Management platform for flight test of agriculture and forestry spraying operation of large unmanned aerial vehicle

Abstract

The invention discloses a management platform for a flight test of agriculture and forestry spraying operation of a large unmanned aerial vehicle, and belongs to the field of unmanned aerial vehicle management platforms, and the management platform comprises a data collection module which is used for collecting flight state data of the unmanned aerial vehicle in real time according to a preset rule; the data analysis module is in communication connection with the data acquisition module, and is used for judging a jitter level based on the machine body jitter data in combination with a preset threshold value, and performing correlation analysis on the influence of jitter on a pesticide box solution unbalance state, an electric quantity abnormal consumption state and spraying uniformity deviation; the decision control module is used for executing at least one of the following operations according to the analysis result: adjusting a flight path or a spraying parameter of the unmanned aerial vehicle to compensate for center-of-gravity shift; optimizing a battery distribution strategy or triggering a low-electric-quantity return instruction; and generating a re-spraying path of the non-uniform area and sending the re-spraying path to an unmanned aerial vehicle flight control system. According to the invention, the distribution deviation state of the pesticide liquid in the pesticide box can be monitored in different areas, early warning can be found in time when the center of gravity shifts, and the flight safety is improved.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle (UAV) management platforms, and in particular to a management platform for flight tests of large-scale UAVs for agricultural and forestry spraying operations. Background Art

[0002] In recent years, drones have been widely used in agricultural and forestry plant protection, but their operational stability in complex environments remains challenging. Therefore, flight tests are required before drones are officially deployed for agricultural and forestry spraying operations. During flight tests, strong winds or malfunctions can easily cause the drone to vibrate. While back-office staff monitor the drone's route, angle, speed, and altitude through a management platform, they often overlook the drone's onboard medicine tank. If the liquid distribution within the tank shifts, this can easily lead to weight deviations, potentially shifting the overall center of gravity.

[0003] Existing technology typically uses pressure sensors to detect the total amount of liquid spray, but lacks zone-based monitoring. This results in no early warning of center-of-gravity shifts. In actual operations, uneven liquid distribution can easily cause drones to lose control, increasing the risk of crashes. This presents certain drawbacks. Therefore, those skilled in the art have developed a management platform for flight tests of large-scale UAV agricultural and forestry spraying operations to address the issues raised in the background technology above. Summary of the Invention

[0004] The purpose of the present invention is to provide a management platform for flight tests of large-scale unmanned aerial vehicles for agricultural and forestry spraying operations, which can monitor the distribution deviation status of the liquid medicine in the medicine box in different areas, and timely detect and warn when the center of gravity shifts, thereby improving flight safety and solving the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] The management platform for large-scale UAV agricultural and forestry spraying flight tests includes:

[0007] The data acquisition module is used to collect UAV flight status data in real time according to preset rules, including:

[0008] The accelerometer (used to measure the linear acceleration of the aircraft), the gyroscope (used to obtain three-dimensional attitude angle data, including pitch angle, roll angle, and yaw angle), and the vibration sensor (used to detect the vibration amplitude and frequency) are used to obtain the vibration amplitude, frequency, and three-dimensional attitude angle data of the aircraft.

[0009] The weight distribution and real-time remaining amount of the medicine liquid are monitored through the medicine box weight sensor and multi-zone pressure sensor (deployed at the bottom of the medicine box to detect the medicine liquid distribution pressure in each zone in real time);

[0010] Record power consumption rate and remaining battery life through the battery management system (BMS);

[0011] The spray flow rate and droplet distribution images are collected through a flow sensor (installed in the spray pipe), a droplet size detector (based on the laser scattering principle, used to analyze the droplet diameter distribution) and a camera (equipped with a wide-angle lens, used to capture droplet deposition images);

[0012] The data analysis module is connected to the data acquisition module and is used to:

[0013] Based on the vibration data of the fuselage, the vibration level is determined in combination with the preset threshold, and the impact of the vibration on the imbalance of the solution in the medicine tank, abnormal power consumption, and spraying uniformity deviation is analyzed.

[0014] Calculate the center of gravity offset based on the weight distribution change of the medicine box, and generate an imbalance warning when the offset exceeds the preset threshold;

[0015] Combining the vibration frequency and motor load data, a power consumption prediction model is established to dynamically adjust the remaining battery life.

[0016] The droplet distribution image is analyzed by image recognition algorithm, and the coverage coefficient of variation (CV value) is calculated to evaluate the spray uniformity deviation;

[0017] The decision control module is used to perform at least one of the following operations based on the analysis results:

[0018] Adjust the drone’s flight path or spraying parameters to compensate for the center of gravity shift;

[0019] Optimize battery allocation strategy or trigger low-battery return command;

[0020] Generate the spraying path for uneven areas and send it to the UAV flight control system.

[0021] As a further solution of the present invention: the preset rules of the data acquisition module include:

[0022] (1) When the vibration amplitude exceeds 2g (measured by the accelerometer) or the frequency is greater than 10Hz (obtained by FFT analysis of the vibration sensor), the high-density data acquisition mode is triggered and the sampling frequency is increased from 1Hz to 50Hz;

[0023] (2) When the weight distribution deviation of the medicine box (calculated by the pressure difference of each partition through the multi-partition pressure sensor data) exceeds 5% for 10 seconds, the multi-sensor cross-validation is started;

[0024] (3) During the spraying operation, the camera captures the droplet deposition image at a frequency of 2 frames per second and stores it in conjunction with the GPS coordinates.

[0025] As a further solution of the present invention: the specific calculation process of the weight distribution deviation of the medicine box is as follows:

[0026] S1: Divide the bottom of the medicine box into N equal-area partitions. Deploy a pressure sensor in each partition to measure the real-time pressure value Pi, where i = 1···N (unit: Pa).

[0027] S2: Calculate the weight of the liquid in each partition Wi = Pi·Si according to the partition area Si;

[0028] The total weight of the liquid is

[0029] S3: Assuming the liquid medicine is evenly distributed, the theoretical weight of each partition is:

[0030]

[0031] S4: Calculate the absolute deviation between the actual weight and the theoretical weight of each partition:

[0032] ΔWi = |Wi-W partition|;

[0033] S5: Calculate the sum of absolute deviations for all partitions:

[0034]

[0035] S6: Calculate the weight distribution deviation percentage of the medicine box:

[0036]

[0037] As a further solution of the present invention: the method for analyzing the imbalance state of the medicine box solution includes:

[0038] Based on the multi-zone pressure sensor data of the medicine box, a weight distribution heat map in a two-dimensional plane is constructed;

[0039] A dynamic center of gravity offset model is established using the Kalman filter algorithm (used to fuse the body vibration data and weight distribution data):

[0040]

[0041] Where xk is the coordinate of the center of gravity at time k, uk is the jitter acceleration input, zk is the sensor observation value, wk and vk are the system noise and observation noise respectively, A is the state transfer matrix, B is the control input matrix, and H is the observation matrix;

[0042] Through Kalman filtering, xk and zk are fused to output accurate center of gravity offset;

[0043] If the center of gravity offset exceeds 10% of the medicine box width for three consecutive samples, it is determined to be an unbalanced state.

[0044] As a further solution of the present invention: the method for establishing the power consumption prediction model includes:

[0045] Based on historical data, a linear relationship between the vibration frequency f (unit: Hz), flight speed v (unit: m / s), and motor current I (unit: A) is fitted: I = k1·f + k2·v + b, where k1 and k2 are regression coefficients and b is the bias term.

[0046] According to the real-time current value and battery capacity, the remaining battery life is dynamically updated as Tremaining = Cremaining / Iavg, where Cremaining is the remaining power (unit: mAh) and Iavg is the average current value within the sliding time window.

[0047] As a further solution of the present invention: the method for evaluating the spraying uniformity deviation includes:

[0048] Binarize the droplet distribution image and extract the droplet coverage area;

[0049] The working area is gridded into 1m×1m units, and the droplet density D of each unit is counted. i (Unit: drops / cm 2 );

[0050] Calculate the coefficient of variation of droplet density over the entire area Where σ is the total number of grid cells D i The standard deviation of all grid cells D i The average density of

[0051] If the CV value is greater than 25%, it is judged to be non-uniform.

[0052] As a further solution of the present invention: the decision control module further includes:

[0053] Generate balancing suggestions based on the center of gravity offset, including adjusting the opening of the tank partition valve (controlling the flow direction of the liquid medicine through the solenoid valve) or releasing the ballast block (built-in jettisonable ballast device);

[0054] When the battery consumption rate exceeds the expected rate by 20%, the flight speed will be forced to be reduced to a safe value;

[0055] For grid cells with CV values ​​exceeding the standard, local spraying paths are generated and nozzle flow is optimized.

[0056] As a further solution of the present invention: it also includes:

[0057] Visual interaction module, used to display jitter spectrum, liquid distribution heat map and power consumption curve in real time;

[0058] The data storage module stores flight data by test batch and supports multi-dimensional retrieval based on time, area or fault type.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] 1. This application solves the systemic problem caused by the vibration of the fuselage in traditional UAV spraying operations through the technical closed loop of "precise data collection-multi-physical field coupling analysis-dynamic closed-loop control", and achieves the "high precision, high reliability, low risk, and low cost" agricultural and forestry plant protection goals, providing key technical support for smart agriculture and green sustainable development.

[0061] 2. This application can monitor the distribution deviation of the liquid medicine within the medicine box by zone and trigger high-density data collection based on preset rules, thereby providing more precise monitoring of the liquid medicine distribution deviation within the medicine box when the aircraft vibrates. Specifically, through the dynamic center of gravity deviation model (Kalman filter), the liquid medicine distribution is monitored in real time, triggering an early warning 10-15 seconds in advance, avoiding flight accidents caused by loss of center of gravity control and reducing the accident rate by 50%.

[0062] 3. This application has a dynamic balancing compensation function, in which the medicine box partition valve adjustment and counterweight release mechanism can restore the center of gravity offset to a safe threshold (<10% of the medicine box width) within 10 seconds, avoiding medicine damage or spray leakage caused by imbalance.

[0063] 4. This application uses a linear regression model based on jitter frequency and flight speed, combined with a sliding window average current to calculate T residual, reducing the endurance prediction error from 20% in traditional models to less than 5%. Furthermore, it automatically reduces flight time or returns to base when power consumption is abnormal, reducing the risk of crashes due to sudden battery depletion and reducing mission interruption rates by 30%.

[0064] 5. This application uses models such as Kalman filtering and coefficient of variation (CV value) to dynamically correlate body vibration with liquid imbalance, power consumption and spraying uniformity, achieving full-link monitoring of "vibration-load-energy consumption-effect", and improving fault location efficiency by 60%. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 A block diagram of the management platform for flight tests of large-scale UAV agricultural and forestry spraying operations. DETAILED DESCRIPTION

[0066] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0067] As mentioned in the background technology of this application, the inventors have discovered through research that the existing drone management platform detects the total amount of liquid medicine through pressure sensors, but does not monitor the area, resulting in the inability to issue an early warning when the center of gravity shifts. In actual operations, the uneven distribution of liquid medicine can easily cause the drone to lose control, increasing the risk of crashing, and there are certain defects.

[0068] In order to solve the above-mentioned defects, the present application discloses a management platform for flight tests of large-scale unmanned aerial vehicles for agricultural and forestry spraying operations, which can monitor the distribution deviation status of the liquid medicine in the medicine box in different zones, and timely detect and warn when the center of gravity shifts, thereby improving flight safety.

[0069] The following will describe in detail how the solution of this application solves the above technical problems with reference to the accompanying drawings.

[0070] See also Figure 1 In an embodiment of the present invention, a management platform for flight tests of large-scale UAV agricultural and forestry spraying operations includes:

[0071] The data acquisition module is used to collect UAV flight status data in real time according to preset rules, including:

[0072] The accelerometer (used to measure the linear acceleration of the fuselage), the gyroscope (used to obtain three-dimensional attitude angle data, including pitch angle, roll angle, and yaw angle) and the vibration sensor (used to detect the vibration amplitude and frequency of the fuselage) are used to obtain the vibration amplitude, frequency and three-dimensional attitude angle data of the fuselage;

[0073] The weight distribution and real-time remaining amount of the medicine liquid are monitored through the medicine box weight sensor and multi-zone pressure sensor (deployed at the bottom of the medicine box to detect the medicine liquid distribution pressure in each zone in real time);

[0074] Record power consumption rate and remaining battery life through the battery management system (BMS);

[0075] The spray flow rate and droplet distribution images are collected through a flow sensor (installed in the spray pipe), a droplet size detector (based on the laser scattering principle, used to analyze the droplet diameter distribution) and a camera (equipped with a wide-angle lens, used to capture droplet deposition images);

[0076] The data analysis module is connected to the data acquisition module and is used to:

[0077] Based on the vibration data of the fuselage, the vibration level is determined in combination with the preset threshold, and the impact of the vibration on the imbalance of the solution in the medicine tank, abnormal power consumption, and spraying uniformity deviation is analyzed.

[0078] Calculate the center of gravity offset based on the weight distribution change of the medicine box, and generate an imbalance warning when the offset exceeds the preset threshold;

[0079] Combining the vibration frequency and motor load data, a power consumption prediction model is established to dynamically adjust the remaining battery life.

[0080] The droplet distribution image is analyzed by image recognition algorithm, and the coverage coefficient of variation (CV value) is calculated to evaluate the spray uniformity deviation;

[0081] The decision control module is used to perform at least one of the following operations based on the analysis results:

[0082] Adjust the drone’s flight path or spraying parameters to compensate for the center of gravity shift;

[0083] Optimize battery allocation strategy or trigger low-battery return command;

[0084] Generate the spraying path for uneven areas and send it to the UAV flight control system.

[0085] This application solves the systemic problems caused by the vibration of the fuselage in traditional drone spraying operations through the technical closed loop of "precision data collection-multi-physical field coupling analysis-dynamic closed-loop control", and achieves the "high precision, high reliability, low risk, and low cost" agricultural and forestry plant protection goals, providing key technical support for smart agriculture and green sustainable development.

[0086] In this embodiment, the preset rules of the data acquisition module include: (1) When the vibration amplitude of the fuselage exceeds 2g (measured by the acceleration sensor) or the frequency is greater than 10Hz (obtained by FFT analysis of the vibration sensor), the high-density data acquisition mode is triggered and the sampling frequency is increased from 1Hz to 50Hz; (2) When the weight distribution deviation of the medicine box (the pressure difference of each partition is calculated by multi-partition pressure sensor data) exceeds 5% for 10 seconds, multi-sensor cross-validation is started; (3) During the spraying operation, the camera captures the droplet deposition image at a frequency of 2 frames per second and stores it in conjunction with the GPS coordinates. Based on the preset threshold, high-density data acquisition (such as 50Hz sampling) can be triggered to accurately capture the instantaneous vibration characteristics. Compared with the traditional fixed-frequency sampling (1Hz), the data resolution is improved by 50 times, providing high-fidelity input for real-time analysis.

[0087] In this embodiment, the specific calculation process of the medicine box weight distribution deviation is as follows: S1: Divide the bottom of the medicine box into N equal-area partitions, deploy a pressure sensor in each partition, and measure the real-time pressure value Pi, i = 1···N (unit: Pa); S2: Calculate the weight of the medicine liquid in each partition Wi = Pi·Si based on the partition area Si; the total medicine liquid weight is S3: Assuming the liquid medicine is evenly distributed, the theoretical weight of each partition is: S4: Calculate the absolute deviation between the actual weight and theoretical weight of each partition: ΔWi = |Wi-Wpartition|; S5: Calculate the sum of the absolute deviations of all partitions: S6: Calculate the weight distribution deviation percentage of the medicine box:

[0088] In this embodiment, after the multi-sensor cross-validation is triggered, the body attitude angle data (such as the roll angle θ and the pitch angle φ) are read synchronously to verify whether the sensor reading deviation is caused by the tilt of the drone and to eliminate misjudgment; the correction formula is: Recalculate the corrected distribution deviation percentage

[0089] In this embodiment, the method for analyzing the imbalance state of the medicine box solution includes: constructing a weight distribution heat map in a two-dimensional plane based on the pressure sensor data of multiple partitions of the medicine box; and establishing a dynamic center of gravity offset model using a Kalman filter algorithm (for fusing the body vibration data and the weight distribution data):

[0090]

[0091] Where xk is the center of gravity coordinate at time k, uk is the jitter acceleration input, zk is the sensor observation value, wk and vk are the system noise and observation noise, respectively, A is the state transition matrix, B is the control input matrix, and H is the observation matrix. A Kalman filter fuses xk and zk to output an accurate center of gravity offset. If the center of gravity offset exceeds 10% of the medicine box width for three consecutive samples, an imbalance is determined. A dynamic center of gravity offset model (Kalman filter) monitors the medicine liquid distribution in real time, triggering an early warning 10-15 seconds in advance, preventing flight accidents caused by center of gravity loss and reducing the accident rate by 50%.

[0092] In this embodiment, the method for establishing a power consumption prediction model includes: fitting a linear relationship between jitter frequency f (unit: Hz), flight speed v (unit: m / s), and motor current I (unit: A) based on historical data: I = k1·f + k2·v + b, where k1 and k2 are regression coefficients and b is a bias term; dynamically updating the remaining flight time T remaining = C remaining / I avg based on the real-time current value and battery capacity, where C remaining is the remaining battery capacity (unit: mAh) and I avg is the average current value within a sliding time window. By calculating T remaining based on the linear regression model of jitter frequency and flight speed, combined with the average current within the sliding window, the flight time prediction error is reduced from 20% in traditional models to less than 5%.

[0093] In this embodiment, the evaluation method of spray uniformity deviation includes: performing binarization processing on the droplet distribution image to extract the droplet coverage area; gridding the working area into 1m×1m units, and calculating the droplet density D of each unit. i (Unit: drops / cm 2 ) ; Calculate the coefficient of variation of droplet density in the entire area Where σ is the total number of grid cells D i The standard deviation of all grid cells D i If the CV value is greater than 25%, it is considered uneven. Using models such as Kalman filtering and the coefficient of variation (CV value), the vibration of the fuselage is dynamically correlated with liquid imbalance, power consumption, and spray uniformity. This enables full-link monitoring of "vibration-load-energy consumption-effect," improving fault location efficiency by 60%.

[0094] In this embodiment, the decision-making control module further includes: generating balancing suggestions based on the center of gravity offset, including adjusting the opening of the medicine box partition valve (controlling the direction of liquid medicine flow through solenoid valves) or releasing counterweights (built-in jettisonable counterweight devices); forcibly reducing the flight speed to a safe value when the power consumption rate exceeds the expected rate by 20%; generating local supplementary spray paths and optimizing nozzle flow for grid cells with excessive CV values. Through gridded droplet density analysis (CV value calculation) and supplementary spray path generation, the coefficient of variation (CV value) of spray coverage is reduced from 40% in traditional manual inspections to less than 20%, and the liquid medicine waste rate is reduced by 10%-15%. In addition, the medicine box partition valve adjustment and counterweight release mechanism can restore the center of gravity offset to a safe threshold (<10% of the medicine box width) within 10 seconds, avoiding pesticide damage or missed spraying problems caused by imbalance. Automatically slowing down or returning to home when power consumption is abnormal, reducing the risk of crashes due to sudden power exhaustion and reducing mission interruption rate by 30%. Through dither-related nozzle flow regulation and flight altitude optimization, the proportion of pesticide liquid drifting to non-target areas is reduced from 15% to below 5%, reducing environmental pollution and the risk of pesticide abuse.

[0095] In this embodiment, the specific steps of balancing the suggestion are:

[0096] Step 1: Detect and locate center of gravity offset

[0097] 1.1 Input data:

[0098] Real-time center of gravity coordinates xk = [x, y] T (calculated by Kalman filter model);

[0099] Medicine box partition pressure sensor data, fuselage attitude angle (roll angle θ and pitch angle φ);

[0100] 1.2 Judgment conditions:

[0101] If the center of gravity offset (x0, y0 are the geometric center coordinates of the medicine box), triggering the balancing suggestion;

[0102] Step 2: Analysis of offset direction and degree:

[0103] 2.1 Determination of offset direction:

[0104] Determine the offset direction (such as left / right, front / back) based on the position of the center of gravity coordinate xk relative to the geometric center x0. For example: if x>x0, the center of gravity is offset to the right; if y>y0, the center of gravity is offset to the front.

[0105] 2.2 Offset level classification:

[0106] Mild offset (Δd≤10% of the medicine box width): only adjust the distribution of the medicine liquid;

[0107] Severe deviation (Δd>10% of the medicine box width): the counterweight needs to be released;

[0108] Step 3: Adjust the opening of the medicine box partition valve:

[0109] 3.1: Liquid flow control

[0110] Objective: By adjusting the valve opening, the liquid medicine will flow in the opposite direction of the center of gravity.

[0111] Operation logic:

[0112] If the center of gravity is biased to the right (x>x0), open the left partition valve and close the right valve to promote the replenishment of liquid medicine to the left;

[0113] The valve opening ratio is dynamically adjusted according to the offset:

[0114] Opening percentage = q*Δd (q is the proportional coefficient, such as q = 5% / cm)

[0115] Execute action:

[0116] Send instructions to the solenoid valve controller to adjust the valve opening proportionally;

[0117] Monitor the changes in the center of gravity after adjustment in real time. If the changes do not meet expectations, gradually increase the opening;

[0118] 3.2: Traffic balance verification

[0119] Feedback mechanism:

[0120] Verify whether the distribution of the liquid medicine tends to be balanced through the pressure sensor data;

[0121] If Δd continues to decrease after adjustment, maintain the current valve state;

[0122] If invalid, switch to the counterweight release strategy;

[0123] Step 4: Release the counterweight

[0124] 4.1: Counterweight selection

[0125] Counterweight deployment:

[0126] The medicine box is symmetrically equipped with throwable counterweights (e.g., four groups on the left / right, front / back);

[0127] The mass of each set of counterweights is m (e.g. m = 0.5 kg);

[0128] Release logic:

[0129] If the center of gravity is shifted to the right, release the counterweight on the left (reducing the weight on the left, equivalent to shifting the center of gravity to the left); the release quantity n is calculated based on the offset:

[0130]

[0131] Among them, W total is the total weight of the medicine box, L is the width of the medicine box;

[0132] 4.2: Execution Release

[0133] Safety verification:

[0134] Make sure the drone's altitude is greater than a safe value (e.g., 5 meters) to avoid the risk of falling objects.

[0135] Check the number of remaining weights to avoid over-releasing.

[0136] Action trigger:

[0137] The counterweight is controlled to disengage via electromagnetic lock;

[0138] After release, the center of gravity is recalculated. If it still exceeds the limit, the release process is repeated;

[0139] Step 5: Closed-loop verification and optimization

[0140] Real-time monitoring:

[0141] Update the center of gravity coordinate xk and confirm whether Δd is restored to within the threshold;

[0142] If the standards are not met, optimize the valve response speed or counterweight release strategy based on the airframe vibration data;

[0143] Record and learn:

[0144] Store balancing operation logs (such as valve opening, balancing weight release amount, and adjustment effect);

[0145] The proportional coefficient q and the counterweight calculation formula parameters are optimized through machine learning algorithms.

[0146] In this embodiment, the application also includes: a visual interaction module for real-time display of jitter spectrum graphs, liquid distribution heat maps and power consumption curves; a data storage module for storing flight data by test batches and supporting multi-dimensional retrieval based on time, region or fault type. Through interactive interfaces such as heat maps and jitter spectrum graphs, the operation status is intuitively displayed, the technical threshold of operators is lowered, and the efficiency of human-machine collaboration is improved by 40%. Historical data storage and machine learning algorithm support can adaptively optimize model parameters (such as regression coefficients k1, k2), which is suitable for different aircraft models and operating environments, and the scalability is significantly enhanced.

[0147] The present invention solves the systemic problems caused by the vibration of the fuselage in traditional UAV spraying operations through the technical closed loop of "precision data collection-multi-physics field coupling analysis-dynamic closed-loop control", achieves the "high precision, high reliability, low risk, and low cost" agricultural and forestry plant protection goals, and provides key technical support for smart agriculture and green sustainable development. In addition, the present application can monitor the distribution offset status of the liquid medicine in the medicine box in a partitioned manner, and trigger high-density data collection based on preset rules, and then perform more precise monitoring of the distribution offset of the liquid medicine in the medicine box when the fuselage shakes. Among them, the liquid medicine distribution is monitored in real time through the dynamic center of gravity offset model (Kalman filter), and an early warning is triggered 10-15 seconds in advance to avoid flight accidents caused by loss of center of gravity control, and the accident rate is reduced by 50%.

[0148] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

[0149] The above is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solutions and inventive concepts of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A management platform for large-scale UAV agricultural and forestry spraying flight tests, characterized by: include: The data acquisition module is used to collect UAV flight status data in real time according to preset rules, including: Acquire the vibration amplitude, frequency and three-dimensional attitude angle data of the fuselage through the acceleration sensor, gyroscope and vibration sensor; Monitor the weight distribution and real-time remaining amount of the liquid medicine through the medicine box weight sensor and multi-zone pressure sensor; Record power consumption rate and remaining battery life through the battery management system; The spray flow rate and droplet distribution images are collected through flow sensors, droplet size detectors and cameras; The data analysis module is connected to the data acquisition module and is used to: Based on the vibration data of the fuselage, the vibration level is determined in combination with the preset threshold, and the impact of the vibration on the imbalance of the solution in the medicine tank, abnormal power consumption, and spraying uniformity deviation is analyzed. Calculate the center of gravity offset based on the weight distribution change of the medicine box, and generate an imbalance warning when the offset exceeds the preset threshold; Combining the vibration frequency and motor load data, a power consumption prediction model is established to dynamically adjust the remaining battery life. The droplet distribution image is analyzed by image recognition algorithm, and the coverage coefficient of variation is calculated to evaluate the spray uniformity deviation; The decision control module is used to perform at least one of the following operations based on the analysis results: Adjust the drone’s flight path or spraying parameters to compensate for the center of gravity shift; Optimize battery allocation strategy or trigger low-battery return command; Generate the spraying path for uneven areas and send it to the UAV flight control system.

2. The management platform for flight test of large-scale UAV agricultural and forestry spraying operations according to claim 1 is characterized in that: The preset rules of the data acquisition module include: (1) When the vibration amplitude exceeds 2g or the frequency is greater than 10Hz, the high-density data acquisition mode is triggered and the sampling frequency is increased from 1Hz to 50Hz; (2) When the weight distribution deviation of the medicine box exceeds 5% for 10 seconds, multi-sensor cross-validation is started; (3) During the spraying operation, the camera captures the droplet deposition image at a frequency of 2 frames per second and stores it in conjunction with the GPS coordinates.

3. The management platform for flight test of large-scale UAV agricultural and forestry spraying operations according to claim 2 is characterized in that: The specific calculation process of the medicine box weight distribution deviation is as follows: S1: Divide the bottom of the medicine box into N equal-area partitions. Deploy a pressure sensor in each partition to measure the real-time pressure value Pi, where i = 1···N. S2: Calculate the weight of the liquid in each partition Wi = Pi·Si according to the partition area Si; The total weight of the liquid is S3: Assuming the liquid medicine is evenly distributed, the theoretical weight of each partition is: S4: Calculate the absolute deviation between the actual weight and the theoretical weight of each partition: ΔWi = |Wi-W partition|; S5: Calculate the sum of absolute deviations for all partitions: S6: Calculate the weight distribution deviation percentage of the medicine box:

4. The management platform for flight testing of large-scale UAV agricultural and forestry spraying operations according to claim 3 is characterized in that: The method for analyzing the imbalance state of the medicine box solution includes: Based on the multi-zone pressure sensor data of the medicine box, a weight distribution heat map in a two-dimensional plane is constructed; The dynamic center of gravity offset model is established through the Kalman filter algorithm: Where xk is the coordinate of the center of gravity at time k, uk is the jitter acceleration input, zk is the sensor observation value, wk and vk are the system noise and observation noise respectively, A is the state transfer matrix, B is the control input matrix, and H is the observation matrix; Through Kalman filtering, xk and zk are fused to output accurate center of gravity offset; If the center of gravity offset exceeds 10% of the medicine box width for three consecutive samples, it is determined to be an unbalanced state.

5. The management platform for flight test of large-scale UAV agricultural and forestry spraying operations according to claim 4 is characterized in that: The method for establishing the power consumption prediction model includes: Based on historical data, a linear relationship between the jitter frequency f, flight speed v, and motor current I is fitted: I = k1·f + k2·v + b, where k1 and k2 are regression coefficients and b is the bias term. According to the real-time current value and battery capacity, the remaining endurance time Tremaining = Cremaining / Iavg is dynamically updated, where Cremaining is the remaining power and Iavg is the average current value in the sliding time window.

6. The management platform for flight test of large-scale UAV agricultural and forestry spraying operations according to claim 5 is characterized in that: The evaluation method of the spraying uniformity deviation includes: Binarize the droplet distribution image and extract the droplet coverage area; The working area is gridded into 1m×1m units, and the droplet density D of each unit is counted. i ; Calculate the coefficient of variation of droplet density over the entire area Where σ is the total number of grid cells D i The standard deviation of all grid cells D i The average density of If the CV value is greater than 25%, it is judged to be non-uniform.

7. The management platform for flight test of large-scale UAV agricultural and forestry spraying operations according to claim 6 is characterized in that: The decision control module further includes: Generate balancing suggestions based on center of gravity offset, including adjusting the opening of the medicine box partition valve or releasing the ballast block; When the battery consumption rate exceeds the expected rate by 20%, the flight speed will be forced to be reduced to a safe value; For grid cells with CV values ​​exceeding the standard, local spraying paths are generated and nozzle flow is optimized.

8. The management platform for flight test of large-scale UAV agricultural and forestry spraying operations according to claim 7 is characterized in that: Also includes: Visual interaction module, used to display jitter spectrum, liquid distribution heat map and power consumption curve in real time; The data storage unit stores flight data by test batch and supports multi-dimensional retrieval based on time, area or fault type.

Images