Large unmanned aerial vehicle agriculture and forestry spraying system

Abstract

The invention, which belongs to the technical field of agricultural aviation, discloses a large unmanned aerial vehicle agriculture and forestry spraying system comprising an image acquisition module, an image analysis module, a flight height control module, a spraying amount control module, a dynamic adjustment calculation module, a pesticide box state monitoring module, a wind field sensing module and a path re-planning module. The image acquisition module adopts a combination of a multispectral camera and a laser radar, and acquires agriculture and forestry images and three-dimensional point cloud data on an advancing route in real time; the image analysis module analyzes agriculture and forestry images based on a deep learning neural network, and extracts agriculture and forestry height distribution characteristics and density distribution characteristics; the flight height control module calculates the optimal flight height according to the agriculture and forestry height distribution characteristics, and comprehensively considers the agriculture and forestry height closest to the route and the overall agriculture and forestry height. The device can effectively improve the spraying uniformity, is high in environmental adaptability, and is high in operation safety at the same time.

Description

Technical Field

[0001] This invention relates to the field of agricultural aviation technology, specifically to a large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system. Background Technology

[0002] With the development of modern agriculture, drone-based agricultural and forestry spraying technology has been widely applied. Traditional small rotary-wing drones suffer from problems such as low operating efficiency, small payload, and poor wind resistance in agricultural and forestry spraying operations. While large fixed-wing drones have advantages such as large payload and long range, they face technical challenges in agricultural and forestry spraying operations, including insufficient precision in flight altitude control, poor spray uniformity, and weak environmental adaptability.

[0003] Existing technologies for large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying systems suffer from the following main drawbacks: First, traditional systems use fixed flight altitude and spray volume settings, which cannot adapt to spatial variations in the height and density of agricultural and forestry crops, resulting in uneven spraying effects, with some areas being over-sprayed while others are under-sprayed. Second, during flight altitude adjustments, the swaying of the solution inside the tank makes it difficult to guarantee the UAV's attitude stability, posing a safety hazard.

[0004] Therefore, those skilled in the art have provided large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying systems to address the problems mentioned in the background section. Summary of the Invention

[0005] The purpose of this invention is to provide a large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system that can effectively improve spraying uniformity, has strong environmental adaptability, and is highly safe to operate, thereby solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system includes an image acquisition module, an image analysis module, a flight altitude control module, a spray volume control module, a dynamic adjustment calculation module, a tank status monitoring module, a wind field perception module, and a path replanning module. The image acquisition module uses a combination of a multispectral camera and a lidar to acquire agricultural and forestry images and three-dimensional point cloud data on the forward route in real time. The acquisition frequency is not less than 10Hz and the image resolution is not less than 1920×1080. The image analysis module analyzes agricultural and forestry images based on deep learning neural networks, extracting the height and density distribution features of agricultural and forestry, with a height measurement accuracy of ±0.2 meters and a density measurement accuracy of ±5%. The flight altitude control module calculates the optimal flight altitude based on the altitude distribution characteristics of agriculture and forestry, and comprehensively considers the altitude of agriculture and forestry closest to the route and the overall altitude of agriculture and forestry, and uses a weighted average algorithm to determine the flight altitude setpoint. The spraying volume control module calculates the optimal spraying volume based on the distribution characteristics of agricultural and forest density, comprehensively considers the agricultural and forest density closest to the route and the overall agricultural and forest density, and uses a fuzzy control algorithm to determine the spraying volume set value. The dynamic adjustment calculation module calculates the optimal time and position for altitude adjustment and spray volume adjustment in advance to ensure that the drone tilt angle is kept within a reasonable range and the adjustment response time does not exceed 2 seconds. The medicine tank status monitoring module monitors the solution level in the medicine tank in real time and dynamically adjusts the maximum allowable tilt angle range of the drone based on the solution level, with a monitoring accuracy of ±1%. The wind field sensing module uses a combination of an ultrasonic anemometer and a barometer to measure wind speed and direction in real time. The wind speed measurement range is 0-15 m / s, with an accuracy of ±0.2 m / s. The path replanning module makes a comprehensive decision on whether to replan the flight route based on wind field data, obstacle distribution, and drone battery power, with a decision cycle of no more than 1 second.

[0007] As a further aspect of the present invention: the spraying volume control module calculates the real-time spraying volume using the following formula: Q=K×ρ×V×η, where Q is the real-time spraying volume (ml / s), K is the density adjustment coefficient (0.8-1.2), ρ is the agricultural and forestry density (g / m³), V is the flight speed (m / s), and η is the spraying efficiency coefficient (0.9-0.98). The flight altitude control module calculates the optimal flight altitude using the following formula: H = α × H near +β×H avg +γ×H max Where H is the optimal flight altitude (m), H near H is the closest to the altitude of the agricultural and forestry along the route (m). avg H represents the overall average height (m). max Let α be the maximum height of the region (m), and let α, β, and γ be weighting coefficients, with α+β+γ=1.

[0008] As a further aspect of the present invention: the specific analysis process of the image analysis module includes: A convolutional neural network was used to extract features from agricultural and forestry images. The network structure consisted of 5 convolutional layers and 3 fully connected layers. A model for calculating the height of agricultural and forest canopies was developed based on 3D point cloud data, with a grid resolution of 0.5 m × 0.5 m. The outline of a single agricultural and forestry plant is identified by an image segmentation algorithm, and the number of agricultural and forestry plants per unit area is calculated as a density index. Establish height-density distribution maps with an update frequency of no less than 1 Hz; High-value and density outliers are filtered, and the Kalman filter algorithm is used to ensure data stability.

[0009] As a further aspect of the present invention: the altitude decision logic of the flight altitude control module includes: Real-time calculation of the average height of agriculture and forest within 3 meters of the closest route as H near ; Calculate the average height of agriculture and forestry within a 50m x 50m area of ​​the current work zone as H. avg ; Identify the highest agricultural and forest height within the current work area as H max ; The weighting coefficients are dynamically adjusted according to the type of agriculture and forestry: for fruit trees, α=0.5, β=0.3, γ=0.2; for forest trees, α=0.4, β=0.4, γ=0.2. Output flight altitude setting value with a control accuracy of ±0.5 meters.

[0010] As a further aspect of the present invention: the density adjustment strategy of the spray volume control module includes: The density of agroforestry is divided into three levels: sparse, medium, and dense, with corresponding density adjustment coefficients K of 0.8, 1.0, and 1.2, respectively. The spray volume is automatically adjusted based on the flight speed; for every 1 m / s increase in flight speed, the spray volume increases by 3%. The spray volume is automatically increased when approaching the end of the dense forest to ensure that the pesticide is fully utilized; Real-time calculation of the matching degree between the remaining amount of pesticide solution and the remaining working area, and dynamic adjustment of the spraying uniformity; The spray volume control accuracy is ±2%, and the response time is no more than 1 second.

[0011] As a further aspect of the present invention: the height adjustment timing calculation method of the dynamic adjustment calculation module includes: Predicting attitude changes during altitude adjustment based on UAV dynamics model; The maximum allowable tilt angle is determined by the solution volume in the medicine tank. The larger the solution volume, the smaller the maximum allowable tilt angle. Calculate the optimal starting position for altitude adjustment to ensure that the altitude adjustment is completed during the straight flight segment; A smooth transition algorithm is used to avoid the liquid from shaking due to high abrupt changes; The tilt angle during the adjustment process is always kept below 80% of the allowable range.

[0012] As a further aspect of the present invention: the tilt angle limiting rules of the medicine box status monitoring module include: When the solution volume in the medicine tank is greater than 80% of its capacity, the maximum allowable tilt angle is 15 degrees. When the solution volume in the medicine tank is 50%-80% of its capacity, the maximum allowable tilt angle is 20 degrees. When the solution level in the medicine tank is less than 50% of its capacity, the maximum permissible tilt angle is 25 degrees. Real-time monitoring of the frequency of liquid sloshing; when the sloshing amplitude exceeds the threshold, the allowable tilt angle is automatically reduced. The tilt angle limit value is dynamically adjusted based on the viscosity characteristics of the drug solution.

[0013] As a further aspect of the present invention: the spraying strategy for the endpoint area of ​​the spraying volume control module includes: Calculate the distance from the current location to the work endpoint in real time; When the distance to the finish line is within 100 meters, activate the finish line spraying optimization mode; Calculate the optimal spraying intensity based on the remaining liquid volume and the remaining working area; While ensuring uniform spraying, gradually increase the spraying volume; Within 10 meters of the finish line, the spraying volume should reach 120%-150% of the normal value.

[0014] As a further aspect of the present invention: the wind impact assessment method of the wind field sensing module includes: Real-time measurement of three-dimensional wind speed and direction data; Calculate the distance at which wind affects the drift of spray droplets; Assess the fog droplet coverage under current wind conditions, and trigger an alert when the coverage is below 85%. Calculate wind resistance stability based on the weight of the drone and the amount of liquid. The update frequency is 20Hz, and the data smoothing process uses a moving average filter.

[0015] As a further aspect of the present invention: the decision logic of the path replanning module includes: Taking into account four factors: obstacle distance, difficulty of bypassing, degree of wind influence, and remaining drone battery power; A multi-objective optimization algorithm is used to calculate the overall benefit value of path replanning; Path replanning is triggered when the overall benefit value exceeds the threshold. Obstacle distance weight: 0.3; difficulty of bypassing obstacles weight: 0.2; wind impact weight: 0.3; remaining battery power weight: 0.2. The replanned spray path ensures that the spraying effect loss does not exceed 10%.

[0016] As a further aspect of the present invention, the specific implementation of the path replanning module includes: Establish a real-time path planning engine based on the A* algorithm; Consider terrain undulations and obstacle distribution to generate alternative paths; Evaluate the spraying effectiveness, energy consumption, and time cost of each alternative route; Select the path with the highest overall score as the new task route; The path planning calculation time is no more than 0.5 seconds, and the planning results are smooth and continuous.

[0017] Compared with the prior art, the beneficial effects of the present invention are: This invention, through the integration of a multi-module intelligent system, achieves highly precise and adaptive control for large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying, significantly improving spray uniformity, operational efficiency, and environmental adaptability. The system dynamically adjusts flight altitude and spray volume based on real-time images and sensor data, overcoming the uneven coverage problem caused by fixed parameters in traditional methods. Optimized adjustment timing and path planning reduce pesticide waste and flight risks. Simultaneously, wind field sensing and tank status monitoring enhance the system's stability and safety under complex weather conditions. Overall, this invention reduces operating costs, improves resource utilization, and provides a reliable and efficient spraying solution for modern agriculture. Attached Figure Description

[0018] Figure 1 This is a structural block diagram of a large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] As mentioned in the background section of this application, research has found that traditional drone systems, with their fixed flight altitude and spray volume settings, cannot adapt to spatial variations in the height and density of agricultural and forestry crops, resulting in uneven spraying effects, with some areas being over-sprayed and others under-sprayed. Furthermore, during flight altitude adjustments, the swaying of the solution within the tank makes it difficult to guarantee the drone's attitude stability, posing a safety hazard and demonstrating certain shortcomings.

[0021] To address the aforementioned shortcomings, this application discloses a large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system, which can effectively improve spraying uniformity, has strong environmental adaptability, and offers high operational safety.

[0022] The following will describe in detail, with reference to the accompanying drawings, how the solution of this application solves the above-mentioned technical problems.

[0023] Please see Figure 1In this embodiment of the invention, a large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system includes an image acquisition module, an image analysis module, a flight altitude control module, a spray volume control module, a dynamic adjustment calculation module, a tank status monitoring module, a wind field perception module, and a path replanning module. The image acquisition module uses a combination of a multispectral camera and LiDAR to acquire real-time images of agricultural and forestry along the route and 3D point cloud data, with an acquisition frequency of at least 10Hz and an image resolution of at least 1920×1080. The image analysis module analyzes the agricultural and forestry images based on a deep learning neural network, extracting the height and density distribution characteristics of the agricultural and forestry, with an altitude measurement accuracy of ±0.2 meters and a density measurement accuracy of ±5%. The flight altitude control module calculates the optimal flight altitude based on the height distribution characteristics of the agricultural and forestry, comprehensively considering the height of the agricultural and forestry closest to the route and the overall agricultural and forestry height, and uses a weighted average algorithm to determine the flight altitude. The system integrates several key technologies: an altitude setpoint, a spray volume control module to calculate the optimal spray volume based on the distribution characteristics of agricultural and forestry density, and a fuzzy control algorithm to determine the spray volume setpoint, considering both the density of agricultural and forestry closest to the route and the overall density of agricultural and forestry; a dynamic adjustment calculation module to pre-calculate the optimal time and location for altitude and spray volume adjustments to ensure that the drone's tilt angle remains within a reasonable range, with an adjustment response time of no more than 2 seconds; a tank status monitoring module to monitor the solution volume in the tank in real time and dynamically adjust the maximum allowable tilt angle range of the drone based on the solution volume, with a monitoring accuracy of ±1%; a wind field perception module using a combination of ultrasonic anemometer and barometer to measure wind speed and direction in real time, with a wind speed measurement range of 0-15 m / s and an accuracy of ±0.2 m / s; and a path replanning module to make a comprehensive decision on whether to replan the flight route based on wind field data, obstacle distribution, and drone battery power, with a decision cycle of no more than 1 second. This application achieves comprehensive intelligent control of the agricultural and forestry spraying process by integrating image acquisition, analysis, altitude control, spray volume control, dynamic adjustment calculation, tank status monitoring, wind field perception, and path replanning modules. The image acquisition module acquires real-time agricultural and forestry images and 3D point cloud data, providing basic environmental information for the system. The image analysis module extracts altitude and density features based on deep learning to ensure data accuracy. The flight altitude control module and spray volume control module dynamically adjust flight parameters according to environmental characteristics to improve spray targeting. The dynamic adjustment calculation module optimizes adjustment timing and position to ensure flight stability. The tank status monitoring module limits the tilt angle based on solution volume to enhance safety. The wind field sensing module monitors meteorological conditions and assesses the impact of spraying. The path replanning module integrates multiple factors to make path adjustments, improving environmental adaptability. Overall, the system achieves precise, automated, and safe spraying operations.

[0024] In this embodiment, the spray volume control module calculates the real-time spray volume using the following formula: Q = K × ρ × V × η, where Q is the real-time spray volume (ml / s), K is the density adjustment coefficient (0.8-1.2), ρ is the agroforestry density (g / m³), V is the flight speed (m / s), and η is the spray efficiency coefficient (0.9-0.98); the flight altitude control module calculates the optimal flight altitude using the following formula: H = α × H near +β×H avg +γ×H max Where H is the optimal flight altitude (m), H near H is the closest to the altitude of the agricultural and forestry along the route (m). avg H represents the overall average height (m). max The maximum altitude (m) in the region is given by α, β, and γ, which are weighting coefficients, and α+β+γ=1. By introducing formulas for calculating spray volume and flight altitude, environmental parameters are transformed into quantitative control indicators, ensuring that the calculations of spray volume and flight altitude are based on scientific models. The spray volume formula comprehensively considers density, velocity, and efficiency coefficients to achieve dynamic optimization of the spray volume; the flight altitude formula integrates local and overall altitude information through weighted averaging to adapt to different agricultural and forestry distribution characteristics. These formulas enhance the system's decision-making logic and repeatability, avoid subjective judgment errors, and improve the uniformity and reliability of the spraying effect.

[0025] In this embodiment, the specific analysis process of the image analysis module includes: extracting agricultural and forestry image features using a convolutional neural network (CNN), with a network structure containing 5 convolutional layers and 3 fully connected layers; calculating an agricultural and forestry canopy height model based on 3D point cloud data, with a grid resolution of 0.5 m × 0.5 m; identifying the outline of individual agricultural and forestry trees using an image segmentation algorithm, and calculating the number of trees per unit area as a density index; establishing a height-density distribution map with an update frequency of no less than 1 Hz; and filtering height and density outliers using a Kalman filter algorithm to ensure data stability. The image analysis module employs CNN and point cloud processing technology to achieve accurate extraction of agricultural and forestry height and density features. By analyzing images through a multi-layer network structure and combining point cloud data to generate a canopy height model, the system can accurately identify the outline of individual agricultural and forestry trees and the density per unit area; establishing and updating a height-density distribution map in real time provides continuous environmental status feedback; and using a Kalman filter algorithm to process outliers ensures data stability and reliability. This module provides high-quality data input for subsequent control modules, laying the foundation for precise control.

[0026] In this embodiment, the altitude decision logic of the flight altitude control module includes: calculating in real time the average altitude of agriculture and forestry within 3 meters of the closest route as H. near ; Calculate the average height of agriculture and forestry within a 50m x 50m area of ​​the current work zone as H. avgIdentify the highest agricultural and forestry height within the current work area as H. max The weighting coefficients are dynamically adjusted according to the type of agriculture and forestry: α=0.5, β=0.3, γ=0.2 for fruit trees and α=0.4, β=0.4, γ=0.2 for forest trees. The flight altitude setpoint is output with a control accuracy of ±0.5 meters. The flight altitude control module calculates the closest route altitude, regional average altitude, and maximum altitude in real time, and dynamically adjusts the weighting coefficients based on the type of agriculture and forestry, achieving adaptive decision-making for flight altitude. Differentiated weight settings are used for different crops such as fruit trees and forest trees to ensure that the altitude setpoint considers both local details and overall trends; the control accuracy reaches ±0.5 meters, ensuring the matching degree between the flight altitude and the agricultural and forest canopy, reducing the risk of collision and optimizing the spraying coverage effect.

[0027] In this embodiment, the density adjustment strategy of the spray volume control module includes: dividing the agricultural and forest density into three levels: sparse, medium, and dense, with corresponding density adjustment coefficients K of 0.8, 1.0, and 1.2, respectively; automatically adjusting the spray volume baseline value according to the flight speed, increasing the spray volume by 3% for every 1 m / s increase in flight speed; automatically increasing the spray volume near the dense forest endpoint to ensure full utilization of the pesticide solution; calculating the matching degree between the remaining pesticide solution and the remaining working area in real time, and dynamically adjusting the spray uniformity; achieving a spray volume control accuracy of ±2% and a response time of no more than 1 second. The spray volume control module divides the agricultural and forest density into sparse, medium, and dense levels and associates them with density adjustment coefficients to achieve graded adjustment of the spray volume; automatically adjusting the baseline value in conjunction with flight speed and increasing the spray volume in the endpoint area ensures full utilization and uniform distribution of the pesticide solution. By calculating the matching degree between the remaining pesticide solution and the working area in real time, the system dynamically optimizes the spraying strategy, avoiding pesticide waste or insufficiency, and improving operational efficiency and resource utilization.

[0028] In this embodiment, the altitude adjustment timing calculation method of the dynamic adjustment calculation module includes: predicting attitude changes during altitude adjustment based on the UAV dynamics model; querying the database of maximum allowable tilt angles according to the amount of medicine tank solution (the more solution, the smaller the maximum allowable tilt angle); calculating the optimal adjustment starting position to ensure altitude adjustment is completed during a straight flight segment; employing a smooth transition algorithm to avoid sudden altitude changes causing medicine tank sloshing; and keeping the tilt angle below 80% of the allowable range throughout the adjustment process. The dynamic adjustment calculation module predicts attitude changes based on the UAV dynamics model and queries the tilt angle database corresponding to the amount of medicine tank solution to determine the optimal timing and position for altitude adjustment. By selecting a straight flight segment for adjustment and employing a smooth transition algorithm, the system avoids sudden altitude changes and medicine tank sloshing; keeping the tilt angle below 80% of the allowable range further ensures flight stability and safety, and guarantees rapid response and reliability during the adjustment process.

[0029] In this embodiment, the tilt angle limitation rules of the medicine tank status monitoring module include: a maximum allowable tilt angle of 15 degrees when the medicine tank solution volume is greater than 80% of its capacity; a maximum allowable tilt angle of 20 degrees when the medicine tank solution volume is between 50% and 80% of its capacity; and a maximum allowable tilt angle of 25 degrees when the medicine tank solution volume is less than 50% of its capacity. The module also monitors the frequency of medicine liquid sloshing in real time, automatically reducing the allowable tilt angle when the sloshing amplitude exceeds a threshold; and dynamically corrects the tilt angle limitation value based on the medicine liquid viscosity characteristics. The medicine tank status monitoring module sets the maximum allowable tilt angle according to solution volume levels and monitors the frequency and amplitude of medicine liquid sloshing in real time, dynamically correcting the tilt angle limitation value. When the solution volume is high, a stricter tilt angle limitation is adopted to reduce the risk of sloshing; adjusting the limitation value according to the medicine liquid viscosity characteristics enhances the system's adaptability and safety. This module provides key constraints for flight control by precisely monitoring the medicine tank status, preventing loss of control due to dynamic changes in the medicine liquid.

[0030] In this embodiment, the spraying strategy for the endpoint area of ​​the spray volume control module includes: real-time calculation of the distance from the current location to the endpoint; activation of the endpoint spraying optimization mode when the distance to the endpoint is within 100 meters; calculation of the optimal spraying intensity based on the remaining pesticide volume and the remaining working area; gradual increase of the spray volume while ensuring spray uniformity; and achieving 120%-150% of the normal spray volume within 10 meters of the endpoint. The spray volume control module activates the optimization mode in the endpoint area, dynamically adjusting the spraying intensity by calculating the distance from the current location to the endpoint and the remaining pesticide volume. By gradually increasing the spray volume while ensuring uniformity, especially reaching 120%-150% of the normal value within 10 meters of the endpoint, the module ensures complete utilization of the pesticide, avoids waste at the end of the operation, and improves the overall economy and effectiveness of the spraying.

[0031] In this embodiment, the wind impact assessment method of the wind field sensing module includes: real-time measurement of three-dimensional wind speed and direction data; calculation of the drift distance of the sprayed droplets; assessment of the droplet coverage rate under the current wind conditions, triggering an early warning when the coverage rate is below 85%; calculation of wind resistance stability based on the drone weight and pesticide volume; an update frequency of 20Hz; and data smoothing using a moving average filter. The wind field sensing module calculates the droplet drift distance and assesses the coverage rate by measuring three-dimensional wind speed and direction in real time, triggering an early warning when the coverage rate is below 85%. By combining the drone weight and pesticide volume to calculate wind resistance stability, the system can comprehensively assess the wind impact; the use of a moving average filter ensures the accuracy and real-time nature of the measurements. This module enhances the system's responsiveness to meteorological changes and ensures effective coverage even under wind interference.

[0032] In this embodiment, the decision logic of the path replanning module includes: comprehensively considering four factors: obstacle distance, detour difficulty coefficient, wind impact, and remaining drone battery power; calculating the comprehensive benefit value of path replanning using a multi-objective optimization algorithm; triggering path replanning when the comprehensive benefit value exceeds a threshold; the obstacle distance weight is 0.3, the detour difficulty weight is 0.2, the wind impact weight is 0.3, and the remaining battery power weight is 0.2; the replanned path ensures that the spraying effect loss does not exceed 10%. The path replanning module comprehensively considers factors such as obstacle distance, detour difficulty, wind impact, and remaining battery power, and uses a multi-objective optimization algorithm to calculate the comprehensive benefit value, triggering replanning when the value exceeds a threshold. By setting weight coefficients, the system prioritizes high-risk factors; the replanned path ensures that the spraying effect loss does not exceed 10%, achieving intelligent obstacle avoidance and path optimization in complex environments, improving the continuity and safety of operations.

[0033] In this embodiment, the specific implementation of the path replanning module includes: establishing a real-time path planning engine based on the A* algorithm; generating alternative paths considering terrain undulations and obstacle distribution; evaluating the spraying effect, energy consumption, and time cost of each alternative path; selecting the path with the highest comprehensive score as the new operation route; the path planning calculation time does not exceed 0.5 seconds, and the planning results are smooth and continuous. The path replanning module generates alternative paths in real time based on the A* algorithm, evaluates the spraying effect, energy consumption, and time cost, and selects the route with the highest comprehensive score. The planning calculation time is controlled within 0.5 seconds, and the results are smooth and continuous, ensuring the feasibility and efficiency of the path. This module optimizes the flight trajectory by quickly responding to environmental changes, reducing operation interruptions, and improving the overall efficiency and adaptability of the system.

[0034] This invention, through the integration of a multi-module intelligent system, achieves highly precise and adaptive control for large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying, significantly improving spray uniformity, operational efficiency, and environmental adaptability. The system dynamically adjusts flight altitude and spray volume based on real-time images and sensor data, overcoming the uneven coverage problem caused by fixed parameters in traditional methods. Optimized adjustment timing and path planning reduce pesticide waste and flight risks. Simultaneously, wind field sensing and tank status monitoring enhance the system's stability and safety under complex weather conditions. Overall, this invention reduces operating costs, improves resource utilization, and provides a reliable and efficient spraying solution for modern agriculture.

[0035] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

[0036] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system, characterized in that: It includes an image acquisition module, an image analysis module, a flight altitude control module, a spray volume control module, a dynamic adjustment calculation module, a medicine tank status monitoring module, a wind field sensing module, and a path replanning module; The image acquisition module uses a combination of a multispectral camera and a lidar to acquire real-time images of agriculture and forestry along the route and three-dimensional point cloud data. The image analysis module analyzes agricultural and forestry images based on deep learning neural networks to extract the height distribution features and density distribution features of agricultural and forestry. The flight altitude control module calculates the optimal flight altitude based on the altitude distribution characteristics of agriculture and forestry, and comprehensively considers the altitude of agriculture and forestry closest to the route and the overall altitude of agriculture and forestry, and uses a weighted average algorithm to determine the flight altitude setpoint. The spraying volume control module calculates the optimal spraying volume based on the distribution characteristics of agricultural and forest density, comprehensively considers the agricultural and forest density closest to the route and the overall agricultural and forest density, and uses a fuzzy control algorithm to determine the spraying volume set value. The dynamic adjustment calculation module calculates the optimal time and position for altitude adjustment and spray volume adjustment in advance to ensure that the drone tilt angle is kept within a reasonable range. The medicine tank status monitoring module monitors the solution volume in the medicine tank in real time and dynamically adjusts the maximum allowable tilt angle range of the drone based on the solution volume. The wind field sensing module uses a combination of an ultrasonic anemometer and a barometer to measure wind speed and direction in real time. The path replanning module makes a comprehensive decision on whether to replan the flight route based on wind field data, obstacle distribution, and the drone's battery level.

2. The large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system according to claim 1, characterized in that, The spray volume control module calculates the real-time spray volume using the following formula: Q=K×ρ×V×η, where Q is the real-time spray volume, K is the density adjustment coefficient, ρ is the agricultural and forestry density, V is the flight speed, and η is the spray efficiency coefficient. The flight altitude control module calculates the optimal flight altitude using the following formula: H = α × H near +β×H avg +γ×H max Where H is the optimal flight altitude, H near To best approximate the altitude of the agricultural and forestry areas along the route, H avg H represents the overall average height. max The maximum height of the region is given by α, β, and γ, which are weighting coefficients and α+β+γ=1.

3. The large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system according to claim 2, characterized in that, The specific analysis process of the image analysis module includes: A convolutional neural network was used to extract features from agricultural and forestry images. The network structure consisted of 5 convolutional layers and 3 fully connected layers. A model for calculating the height of agricultural and forest canopies was developed based on 3D point cloud data, with a grid resolution of 0.5 m × 0.5 m. The outline of a single agricultural and forestry plant is identified by an image segmentation algorithm, and the number of agricultural and forestry plants per unit area is calculated as a density index. Establish height-density distribution maps with an update frequency of no less than 1 Hz; High-value and density outliers are filtered, and the Kalman filter algorithm is used to ensure data stability.

4. The large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system according to claim 3, characterized in that, The altitude decision logic of the flight altitude control module includes: Real-time calculation of the average height of agriculture and forest within 3 meters of the closest route as H near ; Calculate the average height of agriculture and forestry within a 50m x 50m area of ​​the current work zone as H. avg ; Identify the highest agricultural and forest height within the current work area as H max ; The weighting coefficients are dynamically adjusted according to the type of agriculture and forestry: for fruit trees, α=0.5, β=0.3, γ=0.2; for forest trees, α=0.4, β=0.4, γ=0.

2. Output flight altitude setting value.

5. The large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system according to claim 4, characterized in that, The density adjustment strategy of the spray volume control module includes: The density of agroforestry is divided into three levels: sparse, medium, and dense, with corresponding density adjustment coefficients K of 0.8, 1.0, and 1.2, respectively. The spray volume is automatically adjusted based on the flight speed; for every 1 m / s increase in flight speed, the spray volume increases by 3%. The spray volume is automatically increased when approaching the end of the dense forest to ensure that the pesticide is fully utilized; Real-time calculation of the matching degree between the remaining amount of pesticide solution and the remaining working area, and dynamic adjustment of the spraying uniformity; The spray volume control accuracy is ±2%, and the response time is no more than 1 second.

6. The large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system according to claim 5, characterized in that, The height adjustment timing calculation method of the dynamic adjustment calculation module includes: Predicting attitude changes during altitude adjustment based on UAV dynamics model; The maximum allowable tilt angle is determined by the solution volume in the medicine tank. The larger the solution volume, the smaller the maximum allowable tilt angle. Calculate the optimal starting position for altitude adjustment to ensure that the altitude adjustment is completed during the straight flight segment; A smooth transition algorithm is used to avoid the liquid from shaking due to high abrupt changes; The tilt angle during the adjustment process is always kept below 80% of the allowable range.

7. The large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system according to claim 6, characterized in that, The tilt angle limitation rules of the medicine box status monitoring module include: When the solution volume in the medicine tank is greater than 80% of its capacity, the maximum allowable tilt angle is 15 degrees. When the solution volume in the medicine tank is 50%-80% of its capacity, the maximum allowable tilt angle is 20 degrees. When the solution level in the medicine tank is less than 50% of its capacity, the maximum permissible tilt angle is 25 degrees. Real-time monitoring of the frequency of liquid sloshing; when the sloshing amplitude exceeds the threshold, the allowable tilt angle is automatically reduced. The tilt angle limit value is dynamically adjusted based on the viscosity characteristics of the drug solution.

8. The large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system according to claim 7, characterized in that, The spraying strategy for the endpoint area of ​​the spray volume control module includes: Calculate the distance from the current location to the work endpoint in real time; When the distance to the finish line is within 100 meters, activate the finish line spraying optimization mode; Calculate the optimal spraying intensity based on the remaining liquid volume and the remaining working area; While ensuring uniform spraying, gradually increase the spraying volume; Within 10 meters of the finish line, the spraying volume should reach 120%-150% of the normal value.

9. The large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system according to claim 8, characterized in that, The wind impact assessment method of the wind field sensing module includes: Real-time measurement of three-dimensional wind speed and direction data; Calculate the distance at which wind affects the drift of spray droplets; Assess the fog droplet coverage under current wind conditions, and trigger an alert when the coverage is below 85%. Calculate wind resistance stability based on the weight of the drone and the amount of liquid. The update frequency is 20Hz, and the data smoothing process uses a moving average filter.

10. The large-scale unmanned aerial vehicle (UAV) agricultural and forestry spraying system according to claim 9, characterized in that, The decision logic of the path replanning module includes: Taking into account four factors: obstacle distance, difficulty of bypassing, degree of wind influence, and remaining drone battery power; A multi-objective optimization algorithm is used to calculate the overall benefit value of path replanning; Path replanning is triggered when the overall benefit value exceeds the threshold. Obstacle distance weight: 0.3; difficulty of bypassing obstacles weight: 0.2; wind impact weight: 0.3; remaining battery power weight: 0.

2. The replanned spray path ensures that the spraying effect loss does not exceed 10%.

Images