A method and device for autonomous feeding of biological control in rice fields

Abstract

This invention discloses an autonomous feeding method and device for biological control in rice paddies, comprising: acquiring real-time flight status data of an agricultural drone and material property parameters of the biological agent to be applied; calculating and generating a basic application strategy based on the flight status data and material property parameters; collecting real-time environmental disturbance characteristics at the discharge port and adaptively compensating the basic application strategy according to the environmental disturbance characteristics to obtain a comprehensive application strategy; and controlling the feeding device to perform the feeding action according to the comprehensive application strategy. Compared with the prior art, this invention can overcome the shortcomings of uneven application caused by the inability to dynamically adjust with changes in flight speed and terrain, pesticide drift or obstruction caused by the swirling airflow under the rotor at the discharge port, and the inability to predict wind pressure fluctuations and overcome mechanical delays to achieve staggered feeding. This achieves precise matching between the feeding rate and the operating status and significantly improves the ability to resist complex aerodynamic interference.

Description

Technical Field

[0001] This invention relates to the field of agricultural plant protection technology, and more specifically, to a method and equipment for autonomous feeding of biological control materials in rice paddies. Background Technology

[0002] China is a major agricultural country globally, and aerial plant protection has become a core force driving the development of modern green agriculture, following traditional manual operations and ground machinery. The complex and ever-changing farmland environment and rotor airflow disturbances pose severe challenges to the quality of drone-based aerial spraying operations. Solving the problems of wind interference and uneven spreading in precise aerial material delivery requires the innovative development of an intelligent material delivery control system with environmental adaptability. Adaptive intelligent material delivery control systems are widely used in high-standard operational scenarios such as modern smart agricultural equipment construction and biological control in rice paddies. Their construction and innovation are of great significance for optimizing the precise allocation of agricultural materials and promoting the green and high-quality transformation of my country's agriculture.

[0003] In the prior art, there is an invention patent application with publication number CN113859546A, which discloses a centrifugal disc dispensing device for drones. It mainly uses a material drop structure to guide the material to the bottom horizontal centrifugal disc, and drives the disc to rotate at high speed by a motor to achieve wide-area material dispensing coverage. However, the prior art still has problems such as the inability of the constant discharge mode to dynamically adjust with flight speed and terrain changes, resulting in uneven application density; the discharge port is easily affected by the swirling air under the rotor, causing the agent to drift or the discharge to be blocked; and the system cannot effectively predict local transient wind pressure and overcome mechanical response delay to achieve accurate staggered dispensing.

[0004] Therefore, how to provide an autonomous feeding method and equipment for biological control in paddy fields that can overcome the shortcomings of uneven application caused by the inability to dynamically adjust the flight speed and terrain changes, pesticide drift or obstruction caused by the swirling airflow under the rotor at the discharge port, and the inability to predict wind pressure fluctuations and overcome mechanical delays to achieve staggered feeding, thereby achieving precise matching between the discharge rate and the operating status, and significantly improving the ability to resist complex aerodynamic interference, has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides an autonomous feeding method for biological control in paddy fields. This method overcomes the shortcomings of uneven application caused by the inability to dynamically adjust the feeding rate according to flight speed and terrain changes, pesticide drift or obstruction caused by the swirling current under the rotor at the discharge port, and the inability to predict wind pressure fluctuations and overcome mechanical delays to achieve staggered feeding. This method achieves precise matching between the feeding rate and the operating status and significantly improves the ability to resist complex aerodynamic interference.

[0006] The first technical solution provided by this invention is as follows: This invention provides an autonomous feeding method for biological control in paddy fields, comprising the following steps: S1 acquiring real-time flight status data of an agricultural drone and material property parameters of the biological agent to be fed; S2 calculating and generating a basic feeding strategy based on the flight status data and material property parameters; S3 collecting real-time environmental disturbance characteristics of the discharge port, adaptively compensating the basic feeding strategy according to the environmental disturbance characteristics to obtain a comprehensive feeding strategy, and controlling the feeding equipment to perform feeding actions according to the comprehensive feeding strategy.

[0007] Furthermore, in a preferred embodiment of the present invention, a basic delivery strategy is calculated and generated based on the flight status data and material attribute parameters, including: parsing the flight status data, obtaining the real-time flight speed, and calculating the difference between the absolute altitude and the relative downward altitude in the flight status data to obtain the field terrain features; collecting the operating width of the agricultural drone, and calculating the real-time area sweep rate based on the operating width and the real-time flight speed; obtaining the target pesticide application density based on the material attribute parameters, and calculating and obtaining the theoretical discharge rate based on the target pesticide application density and the real-time area sweep rate; correcting the theoretical discharge rate based on the field terrain features to obtain the target discharge rate; and converting the target discharge rate into basic electronic control commands for controlling the discharge equipment based on the preset mechanical discharge constant of the dispensing equipment, thereby generating the basic delivery strategy.

[0008] Furthermore, in a preferred embodiment of the present invention, the theoretical discharge rate is corrected based on the field terrain features to obtain the target discharge rate, including: extracting the instantaneous terrain slope angle based on the field terrain features; constructing an area compensation function based on the geometric mapping relationship between the theoretical horizontal operating area of ​​the plant protection drone and the actual physical slope area under the instantaneous terrain slope angle; calculating and obtaining a slope compensation coefficient based on the instantaneous terrain slope angle and the area compensation function, wherein the slope compensation coefficient represents the expansion ratio of the actual physical slope area relative to the theoretical horizontal operating area; and compensating the theoretical discharge rate based on the slope compensation coefficient to obtain the target discharge rate.

[0009] Furthermore, in a preferred embodiment of the present invention, adaptive compensation is performed on the basic dispensing strategy based on the environmental disturbance characteristics to obtain a comprehensive dispensing strategy, including: real-time acquisition of dynamic air pressure time-series signals at the discharge port based on air pressure sensors, and extraction of air pressure trough time series based on the dynamic air pressure time-series signals; the time intervals in the air pressure trough time series where the air pressure amplitude is lower than a preset threshold are designated as low-disturbance windows; the inherent mechanical response delay time of the feeding equipment is obtained, and feedforward phase shift is performed on the basic dispensing strategy based on the air pressure trough time series and in combination with the inherent mechanical response delay time to generate a comprehensive dispensing strategy.

[0010] Furthermore, in a preferred embodiment of the present invention, the time intervals in the pressure trough time series where the pressure amplitude is lower than a preset threshold are designated as low-disturbance windows, including: extracting the relative downward height based on the field topographic features and calculating the airfall duration of the biological agent based on a free-fall kinematic model; traversing all candidate time intervals in the pressure trough time series where the pressure amplitude is lower than the preset threshold and calculating the windless duration of each candidate time interval; and designating the candidate time intervals where the windless duration is greater than or equal to the airfall duration as low-disturbance windows.

[0011] Furthermore, in a preferred embodiment of the present invention, the basic delivery strategy is subjected to feedforward phase shifting to generate a comprehensive delivery strategy, including: acquiring historical pressure trough time series; calculating the signal fluctuation period of the dynamic pressure time series signal based on the historical pressure trough time series; predicting the next target pressure trough based on the signal fluctuation period to obtain the theoretical arrival time; subtracting the inherent mechanical response delay time from the theoretical arrival time to obtain the feedforward trigger time; and updating the basic delivery strategy with the feedforward trigger time to generate a comprehensive delivery strategy.

[0012] Further, in a preferred embodiment of the present invention, predicting the next target pressure trough based on the signal fluctuation period to obtain the theoretical arrival time includes: extracting the most recent N consecutive pressure trough moments from the historical pressure trough time series, calculating the time interval sequence between adjacent pressure trough moments; performing a weighted moving average calculation on the time interval sequence to obtain the steady-state fluctuation period; extracting the last trough moment from the historical pressure trough time series, adding the last trough moment to the steady-state fluctuation period, and calculating the theoretical arrival time of the next target pressure trough.

[0013] The present invention provides a second technical solution as follows: This invention also provides a self-feeding device for biological control in rice paddies, the device comprising: The medicine tank has multiple diversion and discharge holes arranged in a ring array on its bottom wall; A mounting bracket is provided on the top of the pesticide tank and is used to fix the pesticide tank to the underside of the plant protection drone. An electrically controlled drive assembly is fixed at the center of the bottom outer side of the medicine box; A feeding mechanism is disposed inside the medicine tank and is connected to the electronically controlled drive assembly for transmission. A flow guiding component, the top of which is connected to the bottom of the reagent tank, is used to receive the biological agent discharged from the diversion discharge hole and guide it downward; A chassis, which is connected to the bottom end of the flow guiding component, and the chassis is provided with a discharge port; A seeding drive assembly is disposed above the chassis and is used to drive the chassis to rotate.

[0014] Furthermore, in a preferred embodiment of the present invention, the flow guiding component includes a plurality of material conveying columns arranged in an array, each of the plurality of material conveying columns being connected to a corresponding diversion discharge hole, and a windproof gap being formed between adjacent material conveying columns to allow airflow to pass through.

[0015] Furthermore, in a preferred embodiment of the present invention, the feeding device further includes at least two pressure-equalizing air guide pipes, specifically: The pressure equalizing gas guide tube is disposed within the flow guide assembly and is centrally symmetrically distributed based on the central axis of the medicine tank; One end of the pressure equalization air guide pipe penetrates the bottom wall of the medicine tank and the other end penetrates the chassis, used to balance the air pressure at the discharge port.

[0016] This invention provides an autonomous feeding method for biological control in paddy fields. The method includes: S1 acquiring real-time flight status data of an agricultural drone and material property parameters of the biological agent to be fed; S2 calculating and generating a basic feeding strategy based on the flight status data and material property parameters; S3 collecting real-time environmental disturbance characteristics of the discharge port, adaptively compensating the basic feeding strategy according to the environmental disturbance characteristics to obtain a comprehensive feeding strategy, and controlling the feeding equipment to perform feeding actions according to the comprehensive feeding strategy. This invention, by acquiring real-time flight status data of agricultural drones and the material property parameters of the biological agents to be applied, calculates and generates a basic application strategy. It establishes a direct correlation between the drone's actual flight speed, terrain altitude, and the equipment's dispensing speed, enabling real-time adjustment of the dispensing speed according to changes in external flight conditions. This solves the problem of uneven application caused by the inability of the constant dispensing mode in existing technologies to dynamically adjust to changes in flight speed and terrain. After determining the basic dispensing speed, it further collects real-time environmental disturbance characteristics at the dispensing port and adaptively compensates for the execution timing of the basic application strategy based on the actual measured air pressure fluctuation data. This yields a comprehensive application strategy and controls the dispensing equipment's actions by calculating the period of airflow intensity change. By incorporating the mechanical delay time generated by the equipment's hardware operation, a discharge command is sent to the control system in advance, ensuring that the material exits the discharge port precisely during periods of weak external air pressure disturbance. This overcomes the shortcomings of pesticide drift or obstruction caused by the swirling current under the rotor at the discharge port, and solves the problem of not being able to predict wind pressure fluctuations and overcome mechanical delays to achieve staggered discharge. The aforementioned basic discharge rate characteristics generated based on flight status and the timing compensation characteristics based on environmental disturbances combine to enable the equipment to maintain a dynamic and uniform distribution of the total discharge volume along the overall flight trajectory, and to automatically control the material to avoid strong wind pressure and discharge instantaneously at the execution level. Through the coordination of overall discharge volume control and local staggered discharge timing, the actual operational capability of the feeding equipment to resist complex aerodynamic interference is effectively improved. Compared with existing technologies, this invention overcomes the shortcomings of uneven pesticide application caused by the inability to dynamically adjust with flight speed and terrain changes, pesticide drift or obstruction caused by the swirling current under the rotor at the discharge port, and the inability to predict wind pressure fluctuations and overcome mechanical delays to achieve staggered discharge, thereby achieving precise matching between the discharge rate and the operational status. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1A flowchart illustrating the steps of a self-feeding method for biological control in paddy fields, provided in an embodiment of the present invention; Figure 2 A logical framework diagram for obtaining the comprehensive delivery strategy provided in this embodiment of the invention; Figure 3 This is a schematic diagram of the structure of the self-feeding device for biological control in rice paddies provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the material feeding mechanism provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the flow guiding component provided in an embodiment of the present invention; Figure 6 This is a cross-sectional view of the pressure equalization air guide tube provided in an embodiment of the present invention.

[0019] Reference numerals: 1. Agent tank; 11. Diversion discharge hole; 2. Machine body mounting bracket; 3. Electrically controlled drive assembly; 4. Material feeding mechanism; 5. Flow guiding assembly; 51. Material conveying column; 6. Chassis; 61. Discharge port; 7. Spreading drive assembly; 8. Pressure equalization air guide pipe. Detailed Implementation

[0020] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0021] It should be noted that when a component is referred to as being "fixed to" or "set on" another component, it can be directly on or indirectly set on the other component; when a component is referred to as being "connected to" another component, it can be directly connected to or indirectly connected to the other component.

[0022] It should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "first", "second", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention.

[0023] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" or "several" means two or more, unless otherwise explicitly specified.

[0024] It should be noted that the structures, proportions, sizes, etc., shown in the accompanying drawings of this specification are only used to complement the content disclosed in the specification for those skilled in the art to understand and read, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0025] like Figures 1 to 6 As shown in the embodiment of the present invention, an autonomous feeding method for biological control in paddy fields is provided. It can overcome the shortcomings of uneven application caused by the inability to dynamically adjust the flight speed and terrain changes, pesticide drift or obstruction caused by the swirling airflow under the rotor at the discharge port, and the inability to predict wind pressure fluctuations and overcome mechanical delays to achieve staggered feeding. Thus, it can achieve precise matching between the discharge rate and the operating status, and significantly improve the ability to resist complex aerodynamic interference.

[0026] This invention provides an autonomous feeding method for biological control in paddy fields, specifically including: S1 acquiring real-time flight status data of an agricultural drone and material attribute parameters of the biological agent to be fed; S2 calculating and generating a basic feeding strategy based on the flight status data and material attribute parameters; S3 collecting real-time environmental disturbance characteristics of the discharge port, adaptively compensating the basic feeding strategy according to the environmental disturbance characteristics to obtain a comprehensive feeding strategy, and controlling the feeding equipment to perform feeding actions according to the comprehensive feeding strategy. Therefore, the technical solution of this invention, compared with the prior art, involves acquiring real-time flight status data of an agricultural drone and the material property parameters of the biological agent to be applied, calculating and generating a basic application strategy, and establishing a direct correspondence between the drone's actual flight speed, terrain altitude, and the equipment's dispensing speed. This allows the dispensing speed to be adjusted in real time according to changes in external flight conditions, solving the problem of uneven application caused by the inability of the constant dispensing mode in the prior art to dynamically adjust with changes in flight speed and terrain. After determining the basic dispensing speed, the invention further collects real-time environmental disturbance characteristics at the dispensing port, and adaptively compensates for the execution timing of the basic application strategy based on the actual measured air pressure fluctuation data, obtaining a comprehensive application strategy and controlling the dispensing equipment's actions. By calculating the period of airflow intensity change and combining it with the mechanical delay time generated by the operation of the equipment hardware, the discharge command is sent to the control system in advance, so that the material exits the discharge port during the period when the external air pressure disturbance is weak. This overcomes the shortcomings of the discharge port being affected by the swirling airflow under the rotor, which causes the agent to drift or be blocked. It also solves the problem of not being able to predict wind pressure fluctuations and overcome mechanical delays to achieve staggered material discharge. The above-mentioned basic discharge rate characteristics based on flight status and the timing compensation characteristics based on environmental disturbances are combined to enable the equipment to maintain a dynamic and uniform distribution of the total discharge volume on the overall flight trajectory, and to automatically control the material to avoid strong wind pressure and discharge it instantaneously at the execution level. Through the coordination of overall discharge volume control and local staggered material discharge timing, the actual operation capability of the feeding equipment to resist complex aerodynamic interference is effectively improved. Compared with existing technologies, this invention can overcome the shortcomings of uneven drug application caused by the inability to dynamically adjust the speed and terrain changes, drug drift or obstruction caused by the swirling airflow under the rotor at the discharge port, and the inability to predict wind pressure fluctuations and overcome mechanical delays to achieve staggered discharge, thereby achieving precise matching between discharge rate and operating status.

[0027] The following detailed description of the steps and procedures of the self-feeding method for biological control in paddy fields, using specific embodiments, provides a concrete example.

[0028] Specifically, in a specific embodiment of the present invention, flight status data is parsed to obtain real-time flight speed, and the difference between absolute altitude and relative downward altitude in the flight status data is calculated to obtain field terrain features; the operating width of the agricultural drone is collected, and the real-time area sweep rate is calculated based on the operating width and real-time flight speed; the target application density is obtained based on material attribute parameters, and the theoretical discharge rate is calculated and obtained based on the target application density and real-time area sweep rate; the theoretical discharge rate is corrected based on field terrain features to obtain the target discharge rate; and the target discharge rate is converted into basic electronic control commands for controlling the feeding equipment based on the preset mechanical discharge constant of the feeding equipment to generate a basic delivery strategy.

[0029] In a specific embodiment of the present invention, a flight status data message in the form of a set of key-value pairs is first received. The real-time flight speed is obtained by extracting the single-precision floating-point values ​​of the corresponding speed tags by traversing the key-value pairs. Simultaneously, the absolute altitude floating-point value representing the altitude position and the downward relative altitude floating-point value representing the distance to the work surface below are parsed from the key-value pair set. A subtraction operation is performed between the absolute altitude floating-point value and the downward relative altitude floating-point value, and the difference is stored as a one-dimensional floating-point array representing the field terrain features. In the above operational logic, the difference between the absolute altitude and the downward relative altitude physically represents the current... The system calculates the true elevation of the ground surface under the previous coordinates. It then sequentially writes multiple differences from consecutive timestamps into memory, forming a one-dimensional floating-point array of field terrain features. This array serves as a preferred method for specific operational scenarios. By traversing the trend of differences in this one-dimensional floating-point array, the system can identify various complex terrains. For example, when the difference array exhibits a linear monotonically increasing or decreasing trend, it represents a continuous slope scenario in hilly terrain; when the difference array shows a step difference exceeding a preset threshold between adjacent indices, it represents a sudden change in the elevation of terraced fields; and when the difference array shows isolated spike values, it represents a localized, uneven protrusion in the field.Based on this, the system reads the system configuration file to obtain the swath width constant of the current agricultural drone's mission. It then multiplies the swath width constant with the real-time flight speed to obtain the real-time area sweep rate, representing the area covered per unit time. This parameter is temporarily stored in memory as a double-precision floating-point number. After acquiring the sweep data, the system looks up the target application density based on the material properties of the biological agent to be applied, such as particle size and bulk density, and multiplies the target application density with the temporarily stored real-time area sweep rate to calculate and obtain the theoretical discharge rate, representing the required discharge mass per second under ideal level flight conditions. As a preferred method of the above processing, since the field terrain features reflect the continuous undulations of the working surface below, the system extracts the terrain undulation variance based on the field terrain features in a one-dimensional floating-point array. The least squares algorithm is then used to process the pre-collected data containing different undulation variances. The system performs a polynomial fitting optimization operation on the dataset corresponding to the optimal material compensation amount, and finds the extreme point that minimizes the sum of squared residuals in the multidimensional solution space to calculate the continuous dynamic terrain compensation coefficient. The terrain compensation coefficient corresponding to the field terrain characteristics is multiplied with the theoretical material discharge rate to execute the correction logic, thereby obtaining the target material discharge rate that fits the actual three-dimensional terrain surface area. Subsequently, according to the mechanical displacement constant preset by the feeding equipment to characterize the mass of material discharged in a single cycle, the system divides the target material discharge rate by the mechanical displacement constant to obtain the target speed value, which is then converted into the basic electronic control command for controlling the pulse trigger frequency of the feeding equipment, such as the stepper motor. This command is then uniformly encapsulated into structured data to generate the basic delivery strategy. This allows the underlying operating logic to dynamically change the unit time discharge rate according to the speed and terrain undulations when controlling the scattering action, which can effectively solve the problem of uneven distribution of pesticide application density in farmland caused by simply fixing the discharge rate.

[0030] To verify the actual effect of the above-mentioned dynamically generated basic delivery strategy, a comparative test experiment was constructed for the actual dispersing process. The experimental data selection criteria were real telemetry records generated by multiple flights and dispersing in a test field with continuously changing slope. The data came from the exported files of the airborne flight log and the weighing statistics of 100 standard grid receiving trays with a side length of 50 cm evenly distributed on the ground. The experiment used a baseline group with a fixed dispersing speed and an experimental group using the dynamic correction strategy in the above embodiment for comparison and verification. In the preprocessing stage, the underlying array indexing mechanism combined with the data mask comparison algorithm was used to filter out abnormal samples due to hovering abrupt changes during takeoff and landing, and only the steady-state data samples of the level flight operation segment were retained as the effective analysis objects. The experimental steps were as follows: planning two adjacent flights with consistent terrain undulation. A 50-meter-long flight path was used to direct drones of both the baseline and experimental groups to perform seeding operations at randomly varying speeds between 3 and 6 meters per second. After the operation, all ground receiving trays were retrieved, and the material mass in each grid was weighed using a high-precision electronic balance to calculate the coefficient of variation. The experimental results showed that the baseline group had an overall pesticide density variation coefficient as high as 35% during the acceleration and terrain lifting phases due to the lack of area sweep rate and terrain undulation correction. In contrast, the experimental group, by implementing dynamic calculations and issuing basic electronic control commands, successfully ensured that the drop mass closely matched the real-time speed and actual terrain changes, and its pesticide density variation coefficient steadily decreased to below 8.5%. This fully demonstrates that the method of dynamically calculating flight status and material parameters and generating basic delivery strategies can significantly improve the uniformity of farmland material distribution and three-dimensional terrain adaptability.

[0031] Specifically, in a specific embodiment of the present invention, the theoretical discharge rate is corrected based on the field terrain features to obtain the target discharge rate, including: extracting the instantaneous terrain slope angle based on the field terrain features; constructing an area compensation function based on the geometric mapping relationship between the theoretical horizontal operating area of ​​the plant protection drone and the actual physical slope area under the instantaneous terrain slope angle; calculating and obtaining the slope compensation coefficient based on the instantaneous terrain slope angle and the area compensation function, the slope compensation coefficient representing the expansion ratio of the actual physical slope area relative to the theoretical horizontal operating area; and compensating the theoretical discharge rate based on the slope compensation coefficient to obtain the target discharge rate.

[0032] In this embodiment of the invention, the system iterates through adjacent height difference elements in the field terrain features using low-level pointer displacement operations, performs multiplication operations with the corresponding sampling time interval recorded in the low-level register and the real-time flight speed to obtain the horizontal displacement, and calculates the arctangent value of the ratio of height difference to horizontal displacement by calling the arctangent operator in the basic mathematical library, thereby generating and storing instantaneous terrain slope angle data in double-precision floating-point form in memory. In this embodiment, to solve the problem of actual operating area expansion caused by surface tilt, the system constructs an area compensation function at the software logic level. The mapping logic of this function is to divide the constant 1 by the cosine value of the instantaneous terrain slope angle. The system uses the instantaneous terrain slope angle in memory as the input parameter. The area compensation function performs trigonometric function operations and reciprocal operations to calculate the slope compensation coefficient in single-precision floating-point form, representing the expansion ratio of the actual physical slope area relative to the theoretical horizontal working area. Subsequently, the system reads the single-precision floating-point number representing the theoretical discharge rate previously cached in memory, and performs a low-level floating-point multiplication instruction with the slope compensation coefficient just calculated. This dynamically amplifies and compensates the theoretical discharge rate, obtains the target discharge rate, and overwrites it into the data structure of the system control instruction. This enables the underlying operating logic to automatically increase the discharge amount proportionally according to the slope of the working surface when controlling the spreading action. This effectively solves the technical defects of the traditional two-dimensional area algorithm in three-dimensional undulating farmland, which causes sparse or uneven pesticide application density.

[0033] Specifically, in specific embodiments of the present invention, such as Figure 2 As shown, an adaptive compensation is performed on the basic dispensing strategy based on environmental disturbance characteristics to obtain a comprehensive dispensing strategy. This includes: real-time acquisition of dynamic air pressure time-series signals at the discharge port based on air pressure sensors, and extraction of air pressure trough time series based on the dynamic air pressure time-series signals; marking the time intervals in the air pressure trough time series where the air pressure amplitude is lower than a preset threshold as low-disturbance windows; obtaining the inherent mechanical response delay time of the feeding equipment, and performing feedforward phase shift on the basic dispensing strategy based on the air pressure trough time series and the inherent mechanical response delay time to generate a comprehensive dispensing strategy.

[0034] In this embodiment of the invention, the system continuously reads the digital electrical signals returned by the pressure sensor through the underlying analog-to-digital conversion interface, and pushes them into a one-dimensional floating-point array in memory according to the timestamp sequence to form a dynamic pressure time series signal. The system uses a sliding window algorithm to traverse the one-dimensional floating-point array, extracts local minima by comparing the positive and negative reversal features of the slope within adjacent data index intervals, and encapsulates the corresponding timestamps and pressure amplitudes into a pressure trough time series in key-value pair format. In this embodiment, for the preset threshold on which the calibration low-disturbance window depends, in order to overcome the technical defects of traditional fixed thresholds or simple arithmetic averages being unable to adapt to the nonlinearity and random mutation of the swirl height under the rotor, the system pre-reads pure airflow signals from multiple flights in windless environments. An offline dataset of rotor swirling interference characteristics is stored in memory as a one-dimensional array of massive historical air pressure amplitudes. The system employs a kernel density estimation algorithm, selecting a Gaussian kernel function as a smoothing basis. By calculating the optimal bandwidth parameter, the discrete air pressure amplitudes in the one-dimensional array are continuously mapped, generating a continuous probability density distribution sequence in memory. This mapping process effectively filters out statistical errors caused by high-frequency air pressure noise, accurately restoring the true air pressure distribution pattern in complex flow fields. Subsequently, the system performs a first-order difference operation on the probability density distribution sequence to find stationary points with zero derivatives, locating local minima in the probability density curve that represent relatively flat airflow regions. The air pressure amplitude corresponding to these minima is extracted as a baseline value. After obtaining the baseline value, considering that airflow is prone to unpredictable secondary distortions during actual flight, the system further extracts the standard deviation of data samples within the neighborhood of the baseline value to establish an absolutely strict lower tolerance limit. Based on the Gaussian distribution's three sigma criterion, it calculates a value containing three times the standard deviation, which includes a very high probability confidence interval. This three-times-standard-deviation value is then multiplied by a fixed bias ratio (e.g., 50%) pre-written in the configuration file. Finally, the baseline value is subtracted from this product, thus shifting downwards to an extremely stringent anti-disturbance boundary. This boundary is then used as a preset threshold in single-precision floating-point form and permanently resides in memory. By constructing the preset threshold based on kernel density estimation and the three sigma criterion, the system can eliminate disturbances from the underlying mathematical logic. The inducement of transient spurious minima ensures that the subsequently extracted low-disturbance windows are truly aerodynamically stable windless periods, thus ensuring that the material's falling trajectory is not affected by the shear of residual cyclones. The system iterates through the pressure amplitude elements in the pressure trough time series using memory pointers, extracts continuous time nodes with amplitudes less than the strictly preset threshold, and directly marks them as low-disturbance windows. Subsequently, the system reads the underlying device hardware description file to obtain the inherent mechanical response delay time of the stepper motor. This delay time represents the time difference from the system's power control trigger signal to the actual operation of the physical mechanism and the discharge of materials. If no compensation is made, the actual material discharge time will lag behind the marked low-disturbance window, causing the materials to be blown by strong winds again when discharged.Based on the calibrated low-disturbance window start timestamp, a time subtraction operation is performed to deduct the mechanical delay time from the time axis to obtain the feedforward trigger timestamp. Based on this feedforward trigger timestamp, the basic delivery strategy is updated by time-dimensional translation to generate a comprehensive delivery strategy. This ensures that the control system sends out electronic control signals in advance, so that the execution endpoint of the underlying mechanical action precisely matches the extremely narrow pressure trough range, effectively solving the physical engineering problem of agent drift or discharge blockage caused by sudden changes in the downward swirling flow.

[0035] To verify the actual wind resistance and anti-drift effect of the above-mentioned method based on kernel density estimation and three sigma criterion to strictly calculate the preset threshold and implement the feedforward phase shift strategy, a fixed-point spraying comparative experiment was constructed to target the dynamic interference of swirling under the rotor. The test data came from the time-series log files exported in real time by the airborne barometric pressure sensor and the image resolution coordinates of the gridded optical capture matrix deployed on the ground. In the preprocessing stage, the Laida criterion algorithm was used to remove non-rotor periodic abnormal data samples caused by external natural gusts. The experiment used a baseline group with a threshold set according to conventional manual experience and a fixed frequency of spraying, and compared it with the experimental group with a preset threshold generated by the above-mentioned fusion kernel density estimation and statistical downward shift algorithm. The experimental steps were as follows: control the two groups of agricultural drones to maintain a stable hovering state at a height of 3 meters above the ground, start the spraying equipment to operate continuously for 10 seconds, and after the operation, use optical capture to measure the target value. The matrix analysis method was used to analyze the lateral drift trajectory of falling material groups within half a meter of the discharge port and calculate the coefficient of variation. Experimental results showed that the baseline group, due to an excessively high manually set threshold and failure to accurately capture the true boundary of the airflow flat zone, resulted in a large amount of material being discharged during the wind pressure rise period. The average lateral drift distance of the material in the initial stage of leaving the discharge port reached 115 cm. In contrast, the experimental group, through adaptive fitting of the probability density minimum value by the underlying algorithm and strict subtraction of three times the standard deviation, successfully screened out the purest low-disturbance range. Combined with the early triggering mechanism, more than 90% of the material accurately left the discharge port within the millisecond gap in absolute windless conditions. The average lateral drift distance was significantly reduced to less than 12 cm. This fully demonstrates that the method of extracting a preset threshold based on rigorous statistical features and combining it with mechanical delay time for phase compensation can significantly improve the vertical accuracy of the falling trajectory under extreme aerodynamic interference environments.

[0036] Specifically, in this embodiment of the invention, the time intervals in the pressure trough time series where the pressure amplitude is lower than a preset threshold are designated as low-disturbance windows. This includes: extracting the relative downward height based on the field topography and calculating the airfall duration of the biological agent based on a free-fall kinematic model; traversing all candidate time intervals in the pressure trough time series where the pressure amplitude is lower than the preset threshold and calculating the windless duration of each candidate time interval; and designating the candidate time intervals where the windless duration is greater than or equal to the airfall duration as low-disturbance windows.

[0037] In a specific embodiment of the invention, the system accesses previously constructed field terrain features through low-level memory addressing to extract the downward relative height corresponding to the current system timestamp. This height is represented in the low-level data as a single-precision floating-point value representing the vertical distance between the UAV discharge port and the rice canopy directly below. After obtaining the current height, the system directly calculates the descent time of the biological agent in the air based on this height and a free-fall kinematics model. This descent time accurately represents the absolute physical time span required for a single biological agent to travel from the discharge port to landing in the rice field. Subsequently, the system uses low-level pointers to traverse all candidate time interval structures in the pressure trough time series that have been initially truncated and whose pressure amplitude is below a preset threshold. For each candidate time interval, the system performs a timestamp subtraction operation, that is, subtracting the starting timestamp from the ending timestamp inside the structure to calculate the duration of the corresponding trough. The system calculates the duration of windless operation and converts it into a floating-point time difference. Then, it calls underlying logic instructions to compare this duration with the cached air descent duration, performing a greater than or equal to Boolean logic check. If the Boolean return value is true, the system re-masks the candidate time interval structure, officially designating it as a low-disturbance window and embedding it into the data queue of the integrated delivery strategy. For invalid short-duration intervals with a false Boolean return value, the system directly performs memory release operations to remove them. Through this rigorous duration verification mechanism based on the physical motion model, the system can completely prevent the misjudgment of short-duration transient airflow troughs as safe delivery opportunities from the underlying software logic, ensuring that the material is in a windless environment throughout its entire air descent trajectory. This effectively solves the technical defect that frequent short-lived pseudo-troughs during flight cause the agent to still be affected by shear winds in mid-air.

[0038] Specifically, in a specific embodiment of the present invention, a feedforward phase shift is applied to the basic delivery strategy to generate a comprehensive delivery strategy, including: collecting historical pressure trough time series, calculating the signal fluctuation period of the dynamic pressure time series signal based on the historical pressure trough time series, and predicting the next target pressure trough based on the signal fluctuation period to obtain the theoretical arrival time; subtracting the inherent mechanical response delay time from the theoretical arrival time to obtain the feedforward trigger time; and updating the basic delivery strategy with the feedforward trigger time to generate the comprehensive delivery strategy.

[0039] In this embodiment of the invention, the system first obtains the previously recorded historical pressure trough time series through the memory interface, and calculates the signal fluctuation period reflecting the current swirling pressure based on the time difference characteristics of each time node in the series. After obtaining the signal fluctuation period, the system predicts the next target pressure trough based on the period, and obtains the theoretical arrival time of the expected target trough. Subsequently, the system retrieves the pre-stored inherent mechanical response delay time from the underlying configuration file. This parameter is not determined by manual subjective setting, but is obtained by pre-collecting multiple sets of time samples from the level signal issued by the controller to the displacement generated at the physical discharge port under no-load operation conditions. The least squares method is used to perform linear fitting operation on the sample set to find the value that minimizes the sum of squares of the time deviation in the multidimensional parameter space. The extreme point is calculated; the system subtracts the inherent mechanical response delay time from the predicted theoretical arrival time, performs feedforward calculation operation in the time dimension to obtain the feedforward trigger time, and updates the feedforward trigger time to the data structure field of the feeding task through the instruction overriding mechanism to generate a comprehensive feeding strategy; this phase offset processing can make the control signal sending node arrive before the actual trough arrival time, thereby canceling the action lag caused by the electrical characteristics and mechanical inertia of the actuator, ensuring that the physical moment when the material leaves the discharge port under the action of gravity and the instantaneous low-level interval of the external airflow fluctuation are accurately coincided on the time axis, which can solve the technical defect that the discharge action misses the best windless period and causes agent drift due to the physical lag between the control system and the actuator.

[0040] Specifically, in an embodiment of the present invention, predicting the next target pressure trough based on the signal fluctuation period to obtain the theoretical arrival time includes: extracting the most recent N consecutive pressure trough moments from the historical pressure trough time series, calculating the time interval sequence of adjacent pressure trough moments; performing a weighted moving average calculation on the time interval sequence to obtain the steady-state fluctuation period; extracting the last trough moment from the historical pressure trough time series, adding the last trough moment to the steady-state fluctuation period, and calculating the theoretical arrival time of the next target pressure trough.

[0041] In a specific embodiment of this invention, the system first extracts the three to eight most recent consecutive pressure trough moments from the historical pressure trough time series and temporarily stores them in the task sequence. By performing a subtraction operation on two adjacent timestamps, a time interval sequence reflecting the recent airflow fluctuation rhythm is generated. To ensure that the acquired period can eliminate occasional sampling noise and reflect the true steady-state pattern, the system performs a weighted moving average calculation on the time interval sequence. The weight parameter allocation logic is dynamically generated according to the order of the time nodes from farthest to nearest, and an arithmetic series growth algorithm is used to allocate the weight coefficients. Specifically, the system calculates the sorting index value of the current data point in the sequence and divides it by the arithmetic sum of all index values ​​in the sequence. This ratio is used as the weight contribution of each interval sample to ensure... The closer the fluctuation characteristics are to the current moment, the higher the sensitivity to period prediction. By performing multiplication and accumulation operations on the values ​​of each interval and their corresponding dynamic weights, the steady-state fluctuation period that characterizes the current flow field pulsation law is calculated. Subsequently, the system extracts the last trough moment from the historical sequence, and performs addition and offset operations on the floating-point value corresponding to the last trough moment and the calculated steady-state fluctuation period, thereby accurately calculating the theoretical arrival time of the next target pressure trough. This enables the feed control logic to have the ability to predict the future wind field environment in advance, and provides an accurate time reference for the feedforward phase offset of the subsequent actuators. It can effectively solve the technical defects caused by the deviation of the trough law due to changes in rotor speed or external gusts, which leads to inaccurate prediction.

[0042] This invention also provides a self-feeding device for biological control in rice paddies, such as... Figure 3 and Figure 4 As shown, the equipment includes: a pesticide tank 1, with multiple diversion and discharge holes 11 arranged in a circular array on the bottom wall of the pesticide tank 1; a body mounting bracket 2, which is located on the top of the pesticide tank 1 and is used to fix the pesticide tank to the underside of the plant protection drone; an electronically controlled drive assembly 3, which is fixed at the center of the bottom outer side of the pesticide tank; a feeding mechanism 4, which is located inside the pesticide tank 1 and is connected to the electronically controlled drive assembly 3; a flow guiding assembly 5, whose top end is connected to the bottom end of the pesticide tank 1 and is used to receive the biological agent discharged from the diversion and discharge holes 11 and guide it downward; a chassis 6, which is connected to the bottom end of the flow guiding assembly 5 and has a discharge port 61; and a spreading drive assembly 7, which is located above the chassis 6 and is used to drive the chassis 6 to rotate.

[0043] In a specific embodiment of the present invention, the mounting bracket 2 is a gantry structure, with its two ends fixed to the top outer wall of the agent tank 1 by fasteners. The top crossbeam of the mounting bracket 2 is used to rigidly lock with the frame of the plant protection drone to achieve the suspension of the entire feeding device. Multiple diversion and discharge holes 11 arranged in a circular array are provided on the bottom wall of the agent tank 1 to ensure that the granular formulation inside the tank can fall evenly from multiple points under gravity. The electronically controlled drive assembly 3 is fixed to the center of the bottom outer side of the agent tank 1 by a sealed mounting seat. The feeding mechanism 4 is located on the bottom surface inside the agent tank 1. The central hub of the feeding mechanism 4 is sleeved on the power output shaft of the electronically controlled drive assembly 3 extending upwards through the bottom of the tank. The two are connected by a key and a retaining ring at the shaft end for transmission. When the electronically controlled drive assembly 3... During operation, the power output shaft drives the material feeding mechanism 4 to perform a horizontal rotating sweeping action against the bottom wall, forcibly guiding the material to continuously enter each diversion discharge hole 11 to prevent agglomeration and blockage; the flow guiding component 5 has a vertical support structure, and its top end is tightly connected to the bottom end of the agent tank 1, which is used to receive the dispersed material discharged from the multiple diversion discharge holes 11 above and guide it downward smoothly; the chassis 6 is rotatably supported directly below the flow guiding component 5 through the central bearing assembly, and the chassis 6 has radially distributed discharge ports 61 on its surface; the spreading drive component 7 is set on the central axis of the chassis 6 through the support structure, and its motor spindle is coaxially locked with the central rotating shaft of the chassis 6, which is used to drive the chassis 6 to perform high-speed rotation, so that the material falling into the surface of the chassis 6 from the flow guiding component 5 is thrown outward through the discharge port 61 by high-speed centrifugal force.

[0044] Specifically, in specific embodiments of the present invention, such as Figure 5 As shown, the flow guiding component 5 includes several arrayed material conveying columns 51, each of which is connected to a corresponding diversion and discharge hole 11, and a windproof gap is formed between adjacent material conveying columns 51 to allow airflow to pass through.

[0045] In a specific embodiment of the present invention, the flow guiding component 5 is physically manifested as multiple independent conveying columns 51, which are arranged in an array directly below the reagent tank 1. The upper opening of each conveying column 51 is sealed and connected to a corresponding diversion and discharge hole 11 on the bottom wall of the reagent tank 1, so that the biological agent particles overflowing from each discharge hole can enter an independent vertical descent channel without prematurely dispersing in the air. Adjacent conveying columns 51 are not enclosed by a closed structure, but rather a physical clearance is maintained, thus naturally forming within the array. A continuous windbreak gap is formed; when the high-speed downward swirling air generated by the rotor of the agricultural drone hits the feeding equipment, the strong lateral and shear airflow can directly penetrate the entire array of the guide component 5 along the windbreak gap, avoiding the huge wind resistance and violent aerodynamic vibration caused by the wind-catching effect of the large-area closed guide shroud. At the same time, the pesticide particles wrapped inside the conveying column 51 are vertically accelerated in the relatively static air column until they reach the chassis 6, so that the lightweight material is not affected by the crosswind shear in the initial falling stage after leaving the pesticide tank, directly ensuring the verticality of the falling trajectory and the accuracy of the spreading under high dynamic flight.

[0046] Specifically, in specific embodiments of the present invention, such as Figure 6 As shown, the feeding equipment also includes at least two pressure equalization air guide pipes 8. Specifically, the pressure equalization air guide pipes 8 are set in the flow guiding component 5 and are centrally symmetrically distributed based on the central axis of the reagent tank 1. One end of the pressure equalization air guide pipe 8 penetrates the bottom wall of the reagent tank 1 and the other end penetrates the chassis 6 to balance the air pressure at the discharge port 61.

[0047] In a specific embodiment of the present invention, at least two pressure-equalizing air guide pipes 8 are added inside the feeding device. These pressure-equalizing air guide pipes can be made of high-compressive-strength carbon fiber pipes or rigid polyurethane pipes. In terms of spatial layout, these pressure-equalizing air guide pipes 8 are embedded within the physical space of the flow guiding component 5 and are centrally symmetrically distributed with the central axis of the reagent tank 1 as the reference. For example, two pipes are arranged diagonally at 180 degrees or four pipes are arranged in a 90-degree cross array to ensure the overall mechanical dynamic balance of the chassis 6 during high-speed rotation. At the assembly connection interface, the upper end of each pressure-equalizing air guide pipe 8 vertically penetrates the bottom wall of the reagent tank 1 and extends to the upper space inside the reagent tank. A rubber sealing ring is used to lock the penetration at the assembly surface to prevent material leakage from the outer wall of the pipe. Meanwhile, its lower end extends directly downwards and penetrates the solid surface of the chassis 6, so that the pipe opening is directly exposed to the external free flow field below the discharge port 61. When the high-speed downward swirling flow generated by the rotor of the plant protection drone forms a local high-pressure aerodynamic zone below the discharge port 61, the pressure difference airflow can preferentially be guided upwards along the pressure equalization air guide pipe 8 that penetrates the chassis 6 and released into the cavity inside the agent tank 1. Thus, a completely independent air pressure connection loop is constructed between the top of the agent tank 1 and the discharge port 61, which is independent of the material descent channel. This ensures that the air pressure at the discharge port 61 and the internal environment of the agent tank 1 are dynamically balanced in real time, which can eliminate the wind pressure resistance effect caused by the strong pressure airflow rising upwards from the bottom, and ensure that the lightweight granular formulation can maintain a continuous and constant falling rate under pure gravity and mechanical agitation.

[0048] As described above, the present invention relates to an autonomous feeding method for biological control in paddy fields. The method includes: S1 acquiring real-time flight status data of an agricultural drone and material property parameters of the biological agent to be fed; S2 calculating and generating a basic feeding strategy based on the flight status data and material property parameters; S3 collecting real-time environmental disturbance characteristics of the discharge port, adaptively compensating the basic feeding strategy according to the environmental disturbance characteristics to obtain a comprehensive feeding strategy, and controlling the feeding device to perform feeding actions according to the comprehensive feeding strategy. This invention, by acquiring real-time flight status data of agricultural drones and the material property parameters of the biological agents to be applied, calculates and generates a basic application strategy. It establishes a direct correlation between the drone's actual flight speed, terrain altitude, and the equipment's dispensing speed, enabling real-time adjustment of the dispensing speed according to changes in external flight conditions. This solves the problem of uneven application caused by the inability of the constant dispensing mode in existing technologies to dynamically adjust to changes in flight speed and terrain. After determining the basic dispensing speed, it further collects real-time environmental disturbance characteristics at the dispensing port and adaptively compensates for the execution timing of the basic application strategy based on the actual measured air pressure fluctuation data. This yields a comprehensive application strategy and controls the dispensing equipment's actions by calculating the period of airflow intensity change. By incorporating the mechanical delay time generated by the equipment's hardware operation, a discharge command is sent to the control system in advance, ensuring that the material exits the discharge port precisely during periods of weak external air pressure disturbance. This overcomes the shortcomings of pesticide drift or obstruction caused by the swirling current under the rotor at the discharge port, and solves the problem of not being able to predict wind pressure fluctuations and overcome mechanical delays to achieve staggered discharge. The aforementioned basic discharge rate characteristics generated based on flight status and the timing compensation characteristics based on environmental disturbances combine to enable the equipment to maintain a dynamic and uniform distribution of the total discharge volume along the overall flight trajectory, and to automatically control the material to avoid strong wind pressure and discharge instantaneously at the execution level. Through the coordination of overall discharge volume control and local staggered discharge timing, the actual operational capability of the feeding equipment to resist complex aerodynamic interference is effectively improved. Compared with existing technologies, this invention overcomes the shortcomings of uneven pesticide application caused by the inability to dynamically adjust with flight speed and terrain changes, pesticide drift or obstruction caused by the swirling current under the rotor at the discharge port, and the inability to predict wind pressure fluctuations and overcome mechanical delays to achieve staggered discharge, thereby achieving precise matching between the discharge rate and the operational status.

[0049] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for autonomous feeding in rice paddies for biological control, characterized in that, include: S1 acquires real-time flight status data of the agricultural drone and material property parameters of the biological agent to be applied. S2 calculates and generates a basic delivery strategy based on the flight status data and material attribute parameters; S3 collects the environmental disturbance characteristics of the discharge port in real time, and adaptively compensates the basic feeding strategy according to the environmental disturbance characteristics to obtain a comprehensive feeding strategy, and controls the feeding equipment to perform feeding actions according to the comprehensive feeding strategy.

2. The method for autonomously dispensing a biological control agent in a rice field according to claim 1, wherein Based on the flight status data and material attribute parameters, a basic delivery strategy is calculated and generated, including: The flight status data is analyzed to obtain the real-time flight speed, and the difference between the absolute altitude and the relative altitude seen from below in the flight status data is calculated to obtain the terrain features of the field. The operating width of the agricultural drone is collected, and the real-time area sweep rate is calculated based on the operating width and the real-time flight speed. The target application density is obtained based on the material property parameters, and the theoretical discharge rate is calculated and obtained based on the target application density and the real-time area sweep rate. The theoretical discharge rate is corrected based on the field topography to obtain the target discharge rate; Based on the preset mechanical displacement constant of the feeding device, the target feeding rate is converted into basic electrical control commands to control the feeding device, thereby generating the basic feeding strategy.

3. The method for autonomously dispensing a biological control agent in a rice field according to claim 2, wherein The theoretical discharge rate is corrected based on the field topography to obtain the target discharge rate, including: Extract the instantaneous terrain slope angle based on the described field terrain features; Based on the geometric mapping relationship between the theoretical horizontal operating area of ​​the plant protection drone and the actual physical slope area under the instantaneous terrain slope angle, an area compensation function is constructed. Based on the instantaneous terrain slope angle and the area compensation function, the slope compensation coefficient is calculated and obtained. The slope compensation coefficient represents the expansion ratio of the actual physical slope area relative to the theoretical horizontal working area. The theoretical discharge rate is compensated based on the slope compensation coefficient to obtain the target discharge rate.

4. The self-dispensing method for biological control of a rice field according to claim 2, wherein Based on the environmental disturbance characteristics, the basic delivery strategy is adaptively compensated to obtain a comprehensive delivery strategy, including: The dynamic air pressure time series signal at the discharge port is collected in real time based on the air pressure sensor, and the air pressure trough time series is extracted based on the dynamic air pressure time series signal. The time intervals in the pressure trough time series where the pressure amplitude is lower than a preset threshold are designated as low-disturbance windows. The inherent mechanical response delay time of the feeding device is obtained. Based on the pressure trough time series and the inherent mechanical response delay time, the basic feeding strategy is fed forward with phase shift to generate a comprehensive feeding strategy.

5. The method for autonomous feeding of biological control agents in a rice field according to claim 4, wherein The time intervals in the pressure trough time series where the pressure amplitude is below a preset threshold are designated as low-disturbance windows, including: The relative height below is extracted based on the topographic features of the field, and the air fall time of the biological agent is calculated based on the free fall kinematics model. Traverse all candidate time intervals in the time series of pressure troughs where the pressure amplitude is below a preset threshold, and calculate the duration of windless operation in each candidate time interval; The candidate time intervals in which the windless duration is greater than or equal to the duration of the fall from the sky are designated as low-disturbance windows.

6. The method for autonomous feeding of biological control materials in paddy fields according to claim 4, characterized in that, The basic delivery strategy is fed forward with phase shift to generate a comprehensive delivery strategy, including: Historical pressure trough time series are collected, the signal fluctuation period of the dynamic pressure time series signal is calculated based on the historical pressure trough time series, and the next target pressure trough is predicted based on the signal fluctuation period to obtain the theoretical arrival time. Subtract the inherent mechanical response delay time from the theoretical arrival time to obtain the feedforward trigger time; The basic delivery strategy is updated at the feedforward trigger time to generate a comprehensive delivery strategy.

7. The method for autonomous feeding of biological control materials in paddy fields according to claim 6, characterized in that, Based on the signal fluctuation period, the next target pressure trough is predicted to obtain the theoretical arrival time, including: Extract the most recent continuous time series of historical pressure troughs N For each pressure trough moment, calculate the time interval sequence between adjacent pressure trough moments; The steady-state fluctuation period is obtained by performing a weighted moving average calculation on the time interval sequence. Extract the last trough moment from the historical pressure trough time series, add the last trough moment to the steady-state fluctuation period, and calculate the theoretical arrival time of the next target pressure trough.

8. A self-feeding device for biological control in rice paddies, characterized in that, The device includes: The bottom wall of the medicine box (1) is provided with a plurality of diversion and discharge holes (11) arranged in a ring array. The body mounting bracket (2) is set on the top of the medicine box (1) and is used to fix the medicine box to the belly of the plant protection drone. An electronically controlled drive assembly (3) is fixed at the center of the bottom outer side of the medicine box; Material feeding mechanism (4), which is located inside the medicine tank (1) and is connected to the electric control drive assembly (3) for transmission; The top end of the flow guiding component (5) is connected to the bottom end of the medicine tank (1) to receive the biological agent discharged from the diversion discharge hole (11) and guide it downward; The chassis (6) is connected to the bottom end of the flow guiding component (5), and the chassis is provided with a discharge port (61). A seeding drive assembly (7) is disposed above the chassis (6) and is used to drive the chassis (6) to rotate.

9. The self-feeding device for biological control in paddy fields according to claim 8, characterized in that, The flow guiding component (5) includes a plurality of material conveying columns (51) arranged in an array. The plurality of material conveying columns (51) are respectively connected to the corresponding diversion discharge holes (11), and a windproof gap is formed between adjacent material conveying columns (51) to allow airflow to pass through.

10. The self-feeding device for biological control in paddy fields according to claim 8, characterized in that, The feeding device also includes at least two pressure equalizing air pipes (8), specifically: The equalizing gas guide tube (8) is disposed in the flow guide assembly (5) and is centrally symmetrically distributed based on the central axis of the medicine tank (1); One end of the equalizing air pipe (8) passes through the bottom wall of the medicine tank (1) and the other end passes through the chassis (6) to balance the air pressure at the discharge port (61).

Images