A mangrove intelligent planting method and system based on a UAV fleet, an electronic device, and a storage medium

Abstract

The application discloses a kind of based on unmanned aerial vehicle group's mangrove intelligent planting method, system, electronic equipment and storage medium, it is related to ecological restoration technical field, method includes: S10, the environmental data of target planting area is acquired;S20, water power simulation and habitat suitability assessment are carried out based on the environmental data, and generate mangrove planting navigation chart;S30, for the unmanned aerial vehicle group of execution planting task to carry out task allocation and collaborative path planning;S40, according to the planned path execution delivery operation, and in the process of delivery based on environmental disturbance carries out trajectory deviation prediction and dynamic compensation, and based on the deviation of visual recognition landing point carries out feedback correction;S50, utilize unmanned aerial vehicle group to the periodic monitoring of seedling planted, based on monitoring data evaluates seedling survival rate and growth state, and generates reseeding scheme.The beneficial effects of the present application: it can comprehensively improve the efficiency, precision and survival rate of mangrove planting, reduce the monitoring cost of seedling growth.

Description

Technical Field

[0001] This invention relates to the field of ecological restoration technology, and more specifically, to a method, system, electronic device, and storage medium for intelligent mangrove planting based on a drone swarm. Background Technology

[0002] Mangroves are unique forest ecosystems that grow in the intertidal zone of tropical and subtropical coasts, playing vital ecological roles such as windbreak and shoreline stabilization, water purification, biodiversity conservation, and carbon sequestration. However, due to human activities, climate change, and other factors, the global mangrove area is decreasing at a rate of approximately 1% per year, making mangrove ecological restoration a global environmental issue.

[0003] Traditional mangrove planting methods rely primarily on manual labor, requiring workers to wear waterproof gear and traverse muddy intertidal zones. This process is labor-intensive, inefficient, and costly. Furthermore, the complex environment in which mangroves grow, including tidal fluctuations, water erosion, and salinity changes, all affect seedling survival rates, making large-scale, high-precision, and standardized operations difficult to achieve with manual planting.

[0004] In recent years, drone technology has been widely used in agriculture and forestry, such as drone seeding and pesticide spraying. However, applying drones to mangrove planting faces many technical challenges: (1) The marine environment is complex, and wind, waves and tides have a great impact on the flight stability and deployment accuracy of drones; (2) Mangrove planting requires precise positioning and environmental assessment, and a single drone is difficult to complete large-scale operations; (3) After the seedlings are deployed, their growth status needs to be continuously monitored, and traditional manual inspection is costly and time-consuming.

[0005] Therefore, this invention provides a method, system, electronic device and storage medium for mangrove planting based on a drone swarm, which can comprehensively improve the efficiency, accuracy and survival rate of mangrove planting and reduce the monitoring cost of seedling growth. Summary of the Invention

[0006] To overcome the shortcomings of existing technologies, this invention provides a method, system, electronic device, and storage medium for intelligent mangrove planting based on a drone swarm, which can comprehensively improve the efficiency, accuracy, and survival rate of mangrove planting and reduce the monitoring cost of seedling growth.

[0007] The technical solution adopted by this invention to solve its technical problem is: a smart mangrove planting method based on a drone swarm, wherein the improvement is that the smart mangrove planting method based on a drone swarm includes the following steps: S10 utilizes drones and unmanned boats to acquire environmental data of the target planting area during different tidal periods; S20, Based on the environmental data, perform hydrodynamic simulation and habitat suitability assessment to generate a mangrove planting navigation map with suitable planting zones and recommended tree species; S30, based on the mangrove planting navigation map, performs task allocation and collaborative path planning for the drone swarm performing planting tasks, generating the flight path and deployment sequence of each drone; The S40 uses a swarm of drones to carry out deployment operations according to plan. During the deployment process, it performs ballistic deviation prediction and dynamic compensation based on environmental disturbances, and provides feedback correction based on visual recognition of landing point deviation. The S50 uses a swarm of drones to periodically monitor the planted seedlings, assesses the seedling survival rate and growth status based on the monitoring data, and generates a replanting plan based on the seedling survival rate and growth status.

[0008] Furthermore, in step S10, at low tide, a drone equipped with a multispectral sensor and a lidar sensor is used to conduct a low-altitude, water-surface scan of the target planting area to obtain information on the type of tidal flat substrate and the elevation of the tidal flat; and the tidal flat substrate type information is mapped to the Manning roughness coefficient; at high tide, an unmanned surface vessel equipped with an acoustic Doppler current profiler and a multi-parameter water quality probe is used to cruise on the water surface to obtain the water depth, three-dimensional current velocity, salinity, and turbidity of the target planting area.

[0009] Furthermore, in step S20, the specific method for generating a mangrove planting navigation map with suitable planting zones and recommended tree species identified by performing hydrodynamic simulation and habitat suitability assessment based on the environmental data is as follows: S201, input the tidal flat elevation information, Manning roughness coefficient, water depth, three-dimensional flow velocity, salinity, turbidity and tidal period information into the PINN model, use the two-dimensional shallow water equation as the core loss function of the PINN model for joint training, and inversely retrieve the full-domain continuous hourly water surface elevation field and flow velocity field of the target planting area. S202 calculates the effective water depth based on the water surface elevation field and tidal flat elevation information, and generates a flooding duration map; calculates the bottom shear stress based on the velocity field and Manning roughness coefficient, and generates a water flow scour map; generates a salinity zoning map by performing Kriging spatial interpolation based on unmanned vessel navigation data; and generates a bottom sediment turbidity stability map by inverting the spatiotemporal evolution characteristics of turbidity. S203. The flood duration map, water flow scour map, salinity zoning map and bottom sediment turbidity stability map are used as feature layers and input into the multi-criteria evaluation model. The multi-criteria evaluation model performs pixel-by-pixel spatial algebra operations on the feature layers through a weighted linear combination method, calculates and outputs the comprehensive suitability score of each pixel, and generates a suitability comprehensive score map with continuous score surface based on the comprehensive suitability score of each pixel. S204 divides the suitability comprehensive score map into different levels of partition maps according to the score threshold, and superimposes the salinity tolerance threshold of different mangrove species to match recommended tree species for each partition, and finally generates a smart mangrove planting navigation map.

[0010] Furthermore, in step S30, the specific method for assigning tasks and coordinating path planning for the drone swarm performing planting tasks based on the mangrove planting navigation map, and generating the flight path and deployment sequence of each drone, is as follows: S301, extract the coordinates of suitable planting areas from the mangrove planting navigation map as the target point set, and construct the take-off and landing base station, supply point and target point set as a node network for UAV swarm path planning; S302, based on a node network, constructs a multi-UAV path planning model with the goal of minimizing the total flight cost of a UAV swarm. The total flight cost includes flight distance, flight energy consumption, operation time, and environmental risk cost. The constraints of the multi-UAV path planning model include the maximum payload of a single UAV, battery life, maximum range, unique target point allocation, take-off, landing and return, tidal operation time window, safe distance between UAVs, and avoidance of no-fly zones or obstacles. S303 employs an ant colony optimization algorithm to solve the multi-UAV path planning model. By simulating the path search and pheromone update mechanism of artificial ants, it globally optimizes the task allocation scheme that minimizes the total flight cost while satisfying all constraints. Finally, it outputs the target point access sequence, optimal flight path, and deployment sequence for each UAV.

[0011] Furthermore, in step S40, the specific method for predicting and dynamically compensating for ballistic deviation based on environmental disturbances during the delivery process is as follows: The drone's current position, deployment altitude, flight speed, attitude angle, and ambient wind speed and direction are acquired in real time. Combined with seedling mass, windward area, air resistance coefficient, and initial angle of the deployment device, the two-dimensional ballistic deviation during the seedling's descent is calculated using an aerodynamic model. The release compensation parameters are dynamically generated based on the offset. These parameters include the trigger advance for controlling early release, the pitch angle compensation for the release channel to compensate for longitudinal offset, the yaw angle compensation for the release channel to compensate for lateral offset, and the flight speed correction at the moment of release.

[0012] Furthermore, in step S40, the specific method for feedback correction based on visual recognition landing point deviation is as follows: at the moment the seedling enters the water or mud, the UAV-borne visual recognition system captures and identifies the characteristic water splashes or mud splashes stirred up, calculates the geographical coordinates of the actual physical landing point based on the principle of binocular stereo vision, and calculates the landing point deviation value between the actual landing point and the predetermined target point. If the landing point deviation value exceeds the set threshold, the landing point deviation value is used as the feedback correction amount to dynamically update the subsequent seedling deployment compensation parameters.

[0013] Furthermore, in step S50, the specific method for using a swarm of drones to periodically monitor the planted seedlings, assess the seedling survival rate and growth status based on the monitoring data, and generate a replanting plan based on the seedling survival rate and growth status is as follows: S501 utilizes a swarm of drones equipped with multispectral cameras and lidar to periodically patrol and monitor the target planting area, simultaneously acquiring multispectral images and lidar point cloud data of the target planting area. S502 calculates the normalized vegetation index and red edge index based on multispectral imagery, and uses the inverted leaf chlorophyll content as a spectral physiological characteristic to assess the physiological health of seedlings. At the same time, it generates a canopy height model based on lidar point cloud to extract plant height and canopy width as a three-dimensional structural feature to assess seedling growth. By fusing spectral physiological features and three-dimensional structural features, and comparing them with a preset growth time series baseline, it calculates the seedling survival rate and growth status assessment value of each spatial unit. S503 involves gridding the monitoring area, calculating the average survival rate and growth rate of each grid, and selecting grids with substandard average survival rate and growth rate as potential replanting areas. Spatial neighborhood connectivity analysis is then performed on the potential replanting areas to filter out isolated dead grids and determine the final replanting areas. Based on the survival rate gap, growth rate, and grid area, the replanting priority and required replanting quantity for each replanting area are calculated.

[0014] A smart mangrove planting system based on drone swarms, applied to the aforementioned smart mangrove planting method based on drone swarms, is improved by including: The environmental data acquisition module is used to drive drones and unmanned boats to acquire environmental data of the target planting area at different tidal times. The intelligent planning module is used to allocate tasks and plan collaborative paths for a fleet of drones performing planting tasks based on the mangrove planting navigation map, generating the flight paths and deployment sequence of each drone; and it is also used to allocate tasks and plan collaborative paths for a fleet of drones performing planting tasks based on the mangrove planting navigation map, generating the flight paths and deployment sequence of each drone. The precision delivery module is used to drive the drone swarm to perform delivery operations according to the plan. During the delivery process, it performs ballistic deviation prediction and dynamic compensation based on environmental disturbances, and provides feedback correction based on visual recognition of landing point deviation. The monitoring, analysis and decision-making module is used to drive a swarm of drones to periodically monitor the planted seedlings, evaluate the seedling survival rate and growth status based on the monitoring data, and generate a replanting plan based on the seedling survival rate and growth status.

[0015] An electronic device, improved in that it includes at least one processor and at least one memory, wherein, The memory stores computer-readable instructions; The computer-readable instructions are executed by one or more processors, enabling the electronic device to implement the intelligent mangrove planting method based on a drone swarm as described above.

[0016] An improvement of a storage medium having computer-readable instructions stored thereon, wherein the computer-readable instructions are executed by one or more processors to implement a smart mangrove planting method based on a drone swarm as described above.

[0017] The beneficial effects of this invention are as follows: This invention acquires multi-source environmental data of the target planting area through the collaborative use of drones and unmanned surface vessels (USVs) and performs hydrodynamic simulation and habitat suitability assessment. Using a mangrove planting navigation map, it solves the problem of precise positioning and environmental assessment required for mangrove planting, laying the foundation for improved survival rates. By allocating tasks and planning collaborative paths for the drone swarm, it enables parallel operation of the drone swarm, solving the problem of a single drone being unable to complete large-area operations and improving planting efficiency. Through a closed-loop control mechanism that uses ballistic deviation prediction and dynamic compensation based on environmental disturbances and feedback correction based on visual recognition of landing point deviations, it effectively ensures deployment accuracy under complex sea conditions. By utilizing the drone swarm to periodically monitor planted seedlings and generate replanting plans, it replaces traditional manual inspections with automated, periodic aerial inspections, solving the problems of high cost and long cycle of traditional manual inspections. Therefore, this invention can comprehensively improve the efficiency, accuracy, and survival rate of mangrove planting while reducing the monitoring cost of seedling growth. Attached Figure Description

[0018] Figure 1 A flowchart illustrating an intelligent mangrove planting method based on a drone swarm, according to the present invention; Figure 2 This is a block diagram of a mangrove intelligent planting system based on a fleet of unmanned aerial vehicles (UAVs) according to the present invention. Figure 3 This is a hardware structure diagram of an electronic device as an example embodiment; Figure 4 This is a block diagram illustrating an electronic device as an example embodiment. Detailed Implementation

[0019] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0020] The following will clearly and completely describe the concept, specific structure, and technical effects of the present invention in conjunction with embodiments and accompanying drawings, so as to fully understand the purpose, features, and effects of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are all within the scope of protection of the present invention. Furthermore, all connections / linkages involved in the patent do not simply refer to direct contact between components, but rather to the ability to form a better connection structure by adding or reducing connecting accessories according to specific implementation conditions. The various technical features in this invention can be combined interactively without contradicting each other.

[0021] Reference Figure 1 As shown, this invention discloses a smart mangrove planting method based on a drone swarm, which includes the following steps: S10 utilizes drones and unmanned surface vessels (USVs) to acquire environmental data of the target planting area during different tidal periods. Specifically, at low tide, drones equipped with multispectral sensors and lidar sensors conduct low-altitude surface-level scanning of the target planting area to obtain information on the type and elevation of the tidal flat substrate. The tidal flat substrate type information is then mapped to the Manning roughness coefficient. At high tide, USVs equipped with acoustic Doppler current profilers and multi-parameter water quality probes cruise on the water surface to obtain water depth, three-dimensional flow velocity, salinity, and turbidity of the target planting area.

[0022] It should be noted that in this embodiment, environmental data of the target planting area is accurately acquired by using a collaborative operation mode of drones and unmanned vessels during different tidal periods. Specifically, at low tide, when the tide recedes and a large area of ​​mudflats is exposed, drones equipped with multispectral sensors and lidar sensors conduct low-altitude, close-to-the-mudflat surface scanning. The multispectral sensors analyze the differences in spectral reflectance characteristics of different substrates (such as sand, mud, and reefs) to automatically identify and classify the mudflat substrate types. The classification results (such as sand, mud, and gravel) are mapped to the corresponding Manning roughness coefficients according to hydraulic empirical formulas (the Manning roughness coefficient for sand ranges from 0.01 to 0.02, for mud from 0.020 to 0.025, and for gravel / reef from 0.03 to 0.06). This coefficient is used for subsequent water... The dynamic model calculates key parameters of water flow friction resistance. Simultaneously, a lidar sensor, by emitting laser pulses and receiving signals reflected from the tidal flat surface, can penetrate shallow water or directly measure exposed surfaces, generating a high-precision digital elevation model to accurately obtain tidal flat elevation information. At high tide, when the tidal flat is submerged and drones cannot operate effectively, an unmanned survey vessel cruises on the water surface. The acoustic Doppler current profiler on board emits sound waves underwater and receives echoes reflected from suspended particles in the water, allowing for the measurement of three-dimensional flow velocity and direction at different water depths. Simultaneously, a multi-parameter water quality probe onboard continuously monitors and records salinity, turbidity, and depth indicators of the water. This embodiment, through this time-segmented, platform-segmented, and multi-sensor collaborative operation, systematically collects environmental data of the target planting area within a complete tidal cycle, providing a reliable data foundation for subsequent steps.

[0023] S20, Based on the environmental data, perform hydrodynamic simulation and habitat suitability assessment to generate a mangrove planting navigation map that identifies suitable planting zones and recommended tree species; specifically, the method for generating the mangrove planting navigation map that identifies suitable planting zones and recommended tree species based on the environmental data is as follows: S201, input the tidal flat elevation information, Manning roughness coefficient, water depth, three-dimensional flow velocity, salinity, turbidity and tidal period information into the PINN model, use the two-dimensional shallow water equation as the core loss function of the PINN model for joint training, and inversely retrieve the full-domain continuous hourly water surface elevation field and flow velocity field of the target planting area. S202 calculates the effective water depth based on the water surface elevation field and tidal flat elevation information, and generates a flooding duration map; calculates the bottom shear stress based on the velocity field and Manning roughness coefficient, and generates a water flow scour map; generates a salinity zoning map by performing Kriging spatial interpolation based on unmanned vessel navigation data; and generates a bottom sediment turbidity stability map by inverting the spatiotemporal evolution characteristics of turbidity. S203. The flood duration map, water flow scour map, salinity zoning map and bottom sediment turbidity stability map are used as feature layers and input into the multi-criteria evaluation model. The multi-criteria evaluation model performs pixel-by-pixel spatial algebra operations on the feature layers through a weighted linear combination method, calculates and outputs the comprehensive suitability score of each pixel, and generates a suitability comprehensive score map with continuous score surface based on the comprehensive suitability score of each pixel. S204 divides the suitability comprehensive score map into different levels of partition maps according to the score threshold, and superimposes the salinity tolerance threshold of different mangrove species to match recommended tree species for each partition, and finally generates a smart mangrove planting navigation map.

[0024] It should be noted that, in this embodiment, firstly, the acquired tidal flat elevation information, Manning roughness coefficient, measured water depth, three-dimensional flow velocity, salinity, turbidity, and tidal period information are jointly input into the Physical Information Neural Network (PINN) model. This model uses the physical residuals of the two-dimensional shallow water equations as the core loss function for joint optimization training, thereby learning and accurately retrieving the continuously changing water surface elevation and flow velocity fields of the target area in time and space, to achieve accurate simulation of hydrodynamics under complex tidal conditions. The two-dimensional shallow water equations include a water volume continuity equation and a two-dimensional momentum equation. Let... This refers to the water surface elevation. The elevation of the subgrade. For effective water depth ( ), , They are respectively and directional flow velocity, The roughness coefficient is Manning's coefficient. Represents the two-dimensional plane coordinates of the target water area. Let be the time variable, used to represent different moments in the tidal process; then the two-dimensional shallow water equation includes: ; ; ; In the above formula, The PINN model represents gravitational acceleration. As input, with , , For output, the residuals of the above three equations are calculated using automatic differentiation. , , ,in, , , The calculation expressions are as follows: ; ; ; And based on the above residuals , , Construct the total loss function, the expression of which is: In the formula, To monitor losses based on measured water depth and current velocity data from unmanned surface vessels, Loss due to dynamic boundary conditions at tide gauge stations. , , These are the weighting coefficients; This represents the total number of coordinate points sampled within the target spatiotemporal domain; Next, based on the above inversion results, derivative calculations are performed: the effective water depth is obtained by using the difference between the water surface elevation and the tidal flat elevation, and the time when the effective water depth is greater than the inundation threshold within a preset time period is accumulated to generate an inundation duration map; the distribution of bottom shear stress is calculated based on the flow velocity and Manning roughness coefficient according to the fluid dynamics formula, and a water flow scour map is generated by combining the critical start-up conditions of the bottom sediment; at the same time, kriging spatial interpolation is performed on the unmanned vessel navigation data to generate a salinity zoning map, and bottom sediment stability is inverted based on the spatiotemporal evolution characteristics of turbidity to generate a bottom sediment turbidity stability map; thus, the multi-source heterogeneous data are unified into a raster layer that can be quantified and analyzed. Then, the above four images are used as feature layers and input into the Multi-Criterion Evaluation (MCE) model. This model is based on a weighted linear combination method, and sets the weight coefficients of flooding duration, water flow shear force (corresponding to water flow scouring), salinity, and sediment turbidity stability to 0.35, 0.25, 0.20, and 0.20 respectively, according to the importance of the main controlling ecological factors for early survival of mangrove seedlings. Pixel-by-pixel spatial algebra operations are performed on the feature layers to calculate and output the overall suitability score (SI) for each pixel. The formula for calculating the overall suitability score SI is as follows: ; In the above formula, Indicates a pixel as The overall adaptability score; This represents the normalized suitability score for water flow shear force or scour intensity. This indicates the salinity normalization suitability score; This indicates the salinity normalization suitability score; The normalized adaptability score for sediment turbidity stability is represented by the score; each single factor score is normalized to the [0, 1] interval, and the larger the value, the more suitable it is for mangrove seedling planting. The above calculation method is used to perform spatial algebra operations on each pixel and output a corresponding comprehensive suitability score map with continuous scoring surfaces to achieve a quantitative comprehensive evaluation of planting suitability. Finally, the suitability comprehensive score map of each pixel is divided into four different levels of planting zones: "high, medium, low, and unsuitable" according to preset thresholds. Based on this zoning, the salinity tolerance thresholds of different mangrove species are superimposed to automatically match and recommend the most suitable tree species for each suitable zone. Finally, a smart mangrove planting navigation map is generated and output, which marks the planting zones, recommended tree species, and locations of auxiliary engineering measures. This provides a scientific basis for decision-making on site selection and precise tree species matching for subsequent large-scale planting.

[0025] S30, based on the mangrove planting navigation map, task allocation and collaborative path planning are performed for the drone swarm performing the planting task, generating the flight path and deployment sequence of each drone; specifically, the specific method for performing task allocation and collaborative path planning for the drone swarm performing the planting task based on the mangrove planting navigation map, generating the flight path and deployment sequence of each drone, is as follows: S301, extract the coordinates of suitable planting areas from the mangrove planting navigation map as the target point set, and construct the take-off and landing base station, supply point and target point set as a node network for UAV swarm path planning; S302, based on a node network, constructs a multi-UAV path planning model with the goal of minimizing the total flight cost of a UAV swarm. The total flight cost includes flight distance, flight energy consumption, operation time, and environmental risk cost. The constraints of the multi-UAV path planning model include the maximum payload of a single UAV, battery life, maximum range, unique target point allocation, take-off, landing and return, tidal operation time window, safe distance between UAVs, and avoidance of no-fly zones or obstacles. S303 employs an ant colony optimization algorithm to solve the multi-UAV path planning model. By simulating the path search and pheromone update mechanism of artificial ants, it globally optimizes the task allocation scheme that minimizes the total flight cost while satisfying all constraints. Finally, it outputs the target point access sequence, optimal flight path, and deployment sequence for each UAV.

[0026] It should be noted that in this embodiment, firstly, by parsing the mangrove smart planting navigation map, the spatial coordinates of all suitable planting areas in the map are automatically extracted to form a target point set for the operation. The base stations for UAV take-off and landing, the preset supply points, and the target point set are then constructed into a UAV path planning node network to lay the foundation for subsequent optimization modeling. Next, a UAV path planning model is constructed based on this node network. This model aims to minimize the total flight cost of the entire UAV swarm, and its objective function is expressed as follows: In the formula, Represents the set consisting of the target points; A collection of drones representing a drone swarm; Indicates the first Is the drone operated by a node? fly to node ; Indicates the first A drone from the node fly to node The overall cost of flight; Furthermore, the total flight cost comprehensively quantifies flight distance, energy consumption, operation time, and environmental risks. Meanwhile, the UAV path planning model must meet a series of stringent physical and operational constraints, including the maximum payload of a single UAV, battery life, and maximum range, ensuring that each target point is visited only once by one UAV (target point is uniquely assigned), forcing the UAV to depart from the base station and eventually return, operating within the tidal-permitted operation time window, maintaining a safe distance between UAVs, and avoiding no-fly zones and obstacles. Then, the ant colony optimization algorithm is used to solve the path planning model. Specifically, the algorithm simulates each "artificial ant" as a task allocation scheme, using the reciprocal of the flight cost as a heuristic function to guide the ants to eliminate candidate nodes that do not meet the constraints of payload, endurance, time window, safety distance, or obstacle avoidance during the path construction process, based on the "pheromone" concentration on the path and the probability of selecting the next target point according to the heuristic function. After each iteration, the pheromone of the flight segment is updated according to the pheromone evaporation mechanism and the flight cost of the current better path until the preset number of iterations is reached or the optimization result converges. Finally, the target point allocation results, optimal flight path sequence, and delivery sequence corresponding to each UAV are output to guide the UAV swarm to complete the precise delivery of mangrove seedlings or seeds.

[0027] It should also be noted that the UAV swarm adopts a master-slave distributed architecture including a master control node and several operational nodes. Each UAV constructs a point-to-point communication network through the Flight Ad Hoc Network (FANET) protocol. The system acquires real-time data on sudden environmental disturbances (including sudden wind shear, dynamic obstacles, etc.) and the dynamic state parameters of each UAV (including remaining battery SOC, remaining payload, and current pose). When the state parameters or environmental disturbances are detected to deviate from a preset safety threshold, the system uses the current state as the initial condition and employs a rolling time-domain optimization (RHC) strategy to solve for a local optimum within a limited predicted line-of-sight range, dynamically reconstructing and locally correcting the original flight path and task allocation. Specifically, the rolling time-domain optimization strategy refers to the strategy used when the remaining battery SOC, remaining payload, current position, attitude angle, communication link quality, or environmental disturbance data of the UAV deviate from a preset safety threshold. When the safety threshold is reached, the system does not perform a one-time global replanning of all remaining tasks. Instead, it uses the real-time status of each UAV at the current moment as the initial condition, selects the target points to be deployed within the limited predicted line of sight, available UAVs, and environmental constraints, and constructs a local path and task allocation optimization model. The local optimization problem aims to minimize flight energy consumption, flight distance, task delay, safety risk, and path adjustment range within the predicted line of sight, and is constrained by remaining battery power, remaining payload, safety distance, dynamic obstacle avoidance, no-fly zone avoidance, communication connectivity, and tidal operation time window. After the system obtains the locally optimal flight path and task relay scheme, it only executes the first segment of the path or the first set of control commands, and solves again in the next control cycle based on the updated state information, so that the predicted line of sight continuously rolls forward as the task is executed, thereby dynamically correcting the original flight path and task allocation online.

[0028] The S40 uses a swarm of drones to execute deployment operations according to plan. During the deployment process, it predicts and dynamically compensates for trajectory deviations based on environmental disturbances, and provides feedback corrections based on visually recognized landing point deviations. Specifically, the method for predicting and dynamically compensating for trajectory deviations based on environmental disturbances during the deployment process is as follows: The drone's current position, deployment altitude, flight speed, attitude angle, and ambient wind speed and direction are acquired in real time. Combined with seedling mass, windward area, air resistance coefficient, and initial angle of the deployment device, the two-dimensional ballistic deviation during the seedling's descent is calculated using an aerodynamic model. The release compensation parameters are dynamically generated based on the offset. These parameters include the trigger advance for controlling early release, the pitch angle compensation for the release channel to compensate for longitudinal offset, the yaw angle compensation for the release channel to compensate for lateral offset, and the flight speed correction at the moment of release.

[0029] Furthermore, the specific method for feedback correction based on visual recognition landing point deviation is as follows: at the moment the seedling enters the water or mud, the UAV-borne visual recognition system captures and identifies the characteristic water or mud splashes it generates, calculates the geographic coordinates of the actual physical landing point based on the principle of binocular stereo vision, and calculates the landing point deviation value between the actual landing point and the predetermined target point. If the landing point deviation value exceeds the set threshold, the landing point deviation value is used as the feedback correction amount to dynamically update the subsequent seedling deployment compensation parameters.

[0030] It should be noted that in this embodiment, when the UAV arrives at the target point according to the planned path, its onboard flight control system integrates and acquires its own position, altitude, speed, attitude, and real-time environmental wind speed and direction sensed by sensors in real time. It also uses seedling mass, windward area, and air resistance coefficient as fixed parameters, inputting them into the built-in aerodynamic model. This model simplifies the seedling descent process as a projectile motion process influenced by gravity, air resistance, the UAV's initial velocity, and the wind field. It calculates the seedling descent time based on the drop altitude and combines this with the UAV's flight speed and the surrounding environment. By analyzing environmental wind speed, direction, and air resistance parameters, the heading and lateral deviations of the seedlings during their descent are calculated, thus obtaining a two-dimensional ballistic deviation relative to the predetermined target point. Subsequently, based on the two-dimensional ballistic deviation, the airborne flight control system uses inverse kinematics to calculate in real time and generate a set of executable release compensation parameters. These parameters include: a trigger advance to control the early release position, a pitch angle compensation for adjusting the launch angle to compensate for longitudinal deviation, a yaw angle compensation for compensating for lateral wind deflection, and a flight speed correction for fine-tuning the UAV's own speed. These parameters are simultaneously transmitted to the flight control and release execution mechanisms, working together at the moment of release to preemptively offset environmental disturbances. Upon the seedlings' contact with water or mud, the binocular vision system on the bottom of the UAV immediately captures and identifies the characteristic water or mud splashes generated by the impact. Through stereo vision matching and spatial calculation, the precise geographical coordinates of the actual physical landing point are obtained in real time. The system then calculates the deviation between the actual landing point and the predetermined target point. If the deviation exceeds the set threshold, the deviation value is used as a feedback correction amount to dynamically update the aforementioned compensation parameters in subsequent deployment instructions, thereby forming a continuously self-optimizing closed-loop control loop. This ensures that even under uncertain disturbances such as sea surface gusts, the deployment error can be converged within the allowable range through iterative learning, ultimately achieving stable and high-precision automated planting of mangrove seedlings.

[0031] It should also be noted that the drone is equipped with a seedling delivery device, which includes a mechanical gripping mechanism, a delivery channel, and an attitude adjustment mechanism. The mechanical gripping mechanism extracts seedlings from the cargo compartment and transfers them to the delivery channel, which constrains the initial descent trajectory of the seedlings. At the moment of delivery, the airborne control system calculates the descent trajectory deviation based on the real-time wind speed, wind direction, and drone flight speed, and drives the attitude adjustment mechanism to dynamically adjust the spatial launch angle of the delivery channel to counteract the disturbances caused by the environmental wind field and flight inertia, so that the seedlings enter the water in a preset vertical attitude. In the release operation, the seedlings are pre-loaded into a specific cultivation container. The spatial structural features of the cultivation container include: a buoyancy base made of biodegradable material as the main body; a flexible fixing groove set in the center of the buoyancy base to accommodate and buffer the roots of the mangrove seedlings; a counterweight layer set at the bottom of the buoyancy base to lower the overall center of gravity of the cultivation container so that it can maintain an upright, suspended or bottomed posture under the action of gravity and buoyancy after entering the water; and a nutrient slow-release and anti-algae adhesion composite coating covering the outer surface of the buoyancy base.

[0032] S50: A swarm of drones is used to periodically monitor the planted seedlings. Based on the monitoring data, the seedling survival rate and growth status are assessed, and a replanting plan is generated based on the seedling survival rate and growth status. Specifically, the method for periodically monitoring the planted seedlings using a swarm of drones, assessing the seedling survival rate and growth status based on the monitoring data, and generating a replanting plan based on the seedling survival rate and growth status is as follows: S501 utilizes a swarm of drones equipped with multispectral cameras and lidar to periodically patrol and monitor the target planting area, simultaneously acquiring multispectral images and lidar point cloud data of the target planting area. S502 calculates the normalized vegetation index and red edge index based on multispectral imagery, and uses the inverted leaf chlorophyll content as a spectral physiological characteristic to assess the physiological health of seedlings. At the same time, it generates a canopy height model based on lidar point cloud to extract plant height and canopy width as a three-dimensional structural feature to assess seedling growth. By fusing spectral physiological features and three-dimensional structural features, and comparing them with a preset growth time series baseline, it calculates the seedling survival rate and growth status assessment value of each spatial unit. S503 involves gridding the monitoring area, calculating the average survival rate and growth rate of each grid, and selecting grids with substandard average survival rate and growth rate as potential replanting areas. Spatial neighborhood connectivity analysis is then performed on the potential replanting areas to filter out isolated dead grids and determine the final replanting areas. Based on the survival rate gap, growth rate, and grid area, the replanting priority and required replanting quantity for each replanting area are calculated.

[0033] It should be noted that in this embodiment, firstly, the system schedules a fleet of dedicated UAVs equipped with multispectral cameras and lidar to automatically patrol the target planting area according to a preset cycle, simultaneously acquiring multispectral images and high-density lidar point clouds covering the entire area, providing a comprehensive temporal remote sensing data source for growth assessment; then, the acquired data undergoes in-depth processing and feature-level fusion analysis: on the one hand, the normalized vegetation index and red edge index of each pixel are calculated from the multispectral images, and the photosynthetic capacity and physiological stress status of the leaves are quantitatively inverted through these chlorophyll-sensitive spectral indices, forming spectral physiological characteristics reflecting the physiological health status of seedlings; on the other hand, the lidar point clouds are filtered and classified to generate a high-precision canopy height model, from which the plant height and canopy width of individual plants or statistical units are accurately extracted, forming three-dimensional structural features reflecting the growth of seedlings; subsequently, these two types of heterogeneous features are spatially aligned and fused, and then compared with a preset model representing normal growth trajectory. By comparing the time series baseline of the traces, the survival rate and growth status assessment values ​​of seedlings in each assessment unit are comprehensively and quantitatively calculated, achieving accurate diagnosis from appearance to essence. Finally, based on the above quantitative assessment results, intelligent replanting decision-making is initiated. Specifically, the entire monitoring area is first gridded regularly, and the average survival rate and growth rate of each grid are statistically analyzed to initially screen potential replanting grids below the threshold. Next, to avoid resource waste, spatial neighborhood connectivity analysis is performed on potential grids to automatically filter out spatially isolated dead grids caused by accidental factors, preventing drones from making ineffective round trips for single dead seedlings, thereby identifying the truly necessary large-scale final replanting areas. Finally, based on the survival rate gap (the proportion that needs replanting), growth rate (characterizing recovery potential), and grid area of ​​each replanting area, its replanting priority and required replanting quantity are automatically calculated, generating a precise replanting task list that can directly guide the new round of operations, thus completing the entire process from monitoring and assessment to optimization decision-making.

[0034] Reference Figure 2 As shown, this invention also discloses a mangrove intelligent planting system 600 based on a drone swarm, applied to the mangrove intelligent planting method based on a drone swarm as described above, comprising: The environmental data acquisition module 601 is used to drive the drone and the unmanned boat to acquire environmental data of the target planting area at different tidal periods. The intelligent planning module 602 is used to perform task allocation and collaborative path planning for a fleet of drones performing planting tasks based on the mangrove planting navigation map, and to generate the flight path and deployment sequence of each drone; and is also used to perform task allocation and collaborative path planning for a fleet of drones performing planting tasks based on the mangrove planting navigation map, and to generate the flight path and deployment sequence of each drone. The precision delivery module 603 is used to drive the drone swarm to perform delivery operations according to the plan. During the delivery process, it performs ballistic deviation prediction and dynamic compensation based on environmental disturbances, and performs feedback correction based on visual recognition of landing point deviation. The monitoring, analysis and decision-making module 604 is used to drive a fleet of drones to periodically monitor the planted seedlings, evaluate the seedling survival rate and growth status based on the monitoring data, and generate a replanting plan based on the seedling survival rate and growth status.

[0035] It should be noted that the intelligent mangrove planting strategy based on drone swarms provided in the above embodiments is only an example of the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed. That is, the internal structure of the intelligent mangrove planting system 600 based on drone swarms will be divided into different functional modules to complete all or part of the functions described above. Furthermore, the mangrove intelligent planting system 600 based on drone swarms provided in the above embodiments and the mangrove intelligent planting method based on drone swarms belong to the same concept. The specific way each module performs its operation has been described in detail in the method embodiments, and will not be repeated here.

[0036] Figure 3 A schematic diagram of the structure of an electronic device according to an exemplary embodiment is shown.

[0037] It should be noted that this electronic device is merely an example adapted to the present invention and should not be construed as providing any limitation on the scope of use of the present invention. Furthermore, this electronic device should not be interpreted as requiring or depending on having... Figure 3 One or more components of the exemplary electronic device 2000 shown.

[0038] The hardware structure of electronic devices 2000 can vary significantly due to differences in configuration or performance, such as... Figure 3 As shown, the electronic device 2000 includes: a power supply 210, an interface 230, at least one memory 250, and at least one central processing unit (CPU) 270.

[0039] Specifically, power supply 210 is used to provide operating voltage for various hardware devices on electronic device 2000.

[0040] Interface 230 includes at least one wired or wireless network interface 231 for interacting with external devices. Of course, in other examples adapted to this invention, interface 230 may further include at least one serial-to-parallel conversion interface 233, at least one input / output interface 235, and at least one USB interface 237, etc. Figure 3As shown, this does not constitute a specific limitation.

[0041] The memory 250 serves as a carrier for resource storage and can be a read-only memory, random access memory, disk, or optical disk, etc. The resources stored on it include the operating system 251, application programs 253, and data 255, etc., and the storage method can be temporary storage or permanent storage.

[0042] The operating system 251 is used to manage and control the various hardware devices and application programs 253 on the electronic device 2000, so as to enable the central processing unit 270 to perform calculations and processing on the massive data 255 in the memory 250. It can be Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, etc.

[0043] Application 253 is a computer-readable instruction based on operating system 251 that performs at least one specific task, and may include at least one module ( Figure 3 (Not shown), each module can contain computer-readable instructions for electronic device 2000. For example, a mangrove intelligent planting device based on a drone swarm can be considered as application 253 deployed on electronic device 2000.

[0044] Data 255 may be signal information, etc., and is stored in memory 250.

[0045] The central processing unit 270 may include one or more processors and is configured to communicate with the memory 250 via at least one communication bus to read computer-readable instructions stored in the memory 250, thereby enabling the computation and processing of massive amounts of data 255 in the memory 250. For example, a mangrove intelligent planting method based on a drone swarm can be implemented by the central processing unit 270 reading a series of computer-readable instructions stored in the memory 250.

[0046] Furthermore, the present invention can also be implemented through hardware circuits or a combination of hardware circuits and software. Therefore, the implementation of the present invention is not limited to any specific hardware circuit, software, or combination thereof.

[0047] Please see Figure 4 This invention provides an electronic device 4000, which may include: a desktop computer, a laptop computer, a server, etc., with sensor recognition capabilities.

[0048] exist Figure 4 In this context, the electronic device 4000 includes at least one processor 4001 and at least one memory 4003.

[0049] The data interaction between the processor 4001 and the memory 4003 can be achieved through at least one communication bus 4002. This communication bus 4002 may include a path for transmitting data between the processor 4001 and the memory 4003. The communication bus 4002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The communication bus 4002 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0050] Optionally, the electronic device 4000 may further include a transceiver 4004, which can be used for data interaction between the electronic device and other electronic devices, such as sending and / or receiving data. It should be noted that in practical applications, the transceiver 4004 is not limited to one type, and the structure of the electronic device 4000 does not constitute a limitation on the embodiments of the present invention.

[0051] Processor 4001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this invention. Processor 4001 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0052] The memory 4003 may be a ROM (Read Only Memory) or other type of static storage device capable of storing static information and instructions, RAM (Random Access Memory) or other type of dynamic storage device capable of storing information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program instructions or code in the form of instructions or data structures and accessible by the electronic device 4000, but not limited thereto.

[0053] The memory 4003 stores computer-readable instructions, and the processor 4001 can read the computer-readable instructions stored in the memory 4003 through the communication bus 4002.

[0054] The computer-readable instructions are executed by one or more processors 4001 to implement the intelligent mangrove planting method based on drone swarms in the above embodiments.

[0055] Furthermore, this embodiment of the invention provides a storage medium storing computer-readable instructions, which are executed by one or more processors to implement the intelligent mangrove planting method based on a drone swarm as described above.

[0056] Compared with existing technologies, the beneficial effects of this invention are as follows: By collaboratively acquiring multi-source environmental data of the target planting area through UAVs and unmanned vessels and conducting hydrodynamic simulation and habitat suitability assessment, a mangrove planting navigation map is used, solving the problem of precise positioning and environmental assessment required for mangrove planting, laying the foundation for improving survival rates; By allocating tasks and planning collaborative paths for UAV swarms, parallel operation of the UAV swarm is achieved, solving the problem that a single UAV cannot complete large-area operations, thus improving planting efficiency; Through a closed-loop control mechanism that uses ballistic deviation prediction and dynamic compensation based on environmental disturbances and feedback correction based on visual recognition of landing point deviations, the accuracy of deployment under complex sea conditions is effectively guaranteed; By using UAV swarms to periodically monitor planted seedlings and generate replanting plans, automated and periodic aerial inspections replace traditional manual inspections, solving the problems of high cost and long cycle of traditional manual inspections; thereby comprehensively improving the efficiency, accuracy, and survival rate of mangrove planting and reducing the monitoring cost of seedling growth.

[0057] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the embodiments described. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

Claims

1. A method for intelligent mangrove planting based on unmanned aerial vehicle (UAV) swarms, characterized in that, The intelligent mangrove planting method based on drone swarms includes the following steps: S10 utilizes drones and unmanned boats to acquire environmental data of the target planting area during different tidal periods; S20, Based on the environmental data, perform hydrodynamic simulation and habitat suitability assessment to generate a mangrove planting navigation map with suitable planting zones and recommended tree species; S30, based on the mangrove planting navigation map, performs task allocation and collaborative path planning for the drone swarm performing planting tasks, generating the flight path and deployment sequence of each drone; The S40 uses a swarm of drones to carry out deployment operations according to plan. During the deployment process, it performs ballistic deviation prediction and dynamic compensation based on environmental disturbances, and provides feedback correction based on visual recognition of landing point deviation. The S50 uses a swarm of drones to periodically monitor the planted seedlings, assesses the seedling survival rate and growth status based on the monitoring data, and generates a replanting plan based on the seedling survival rate and growth status.

2. The intelligent mangrove planting method based on unmanned aerial vehicle (UAV) swarms according to claim 1, characterized in that, In step S10, at low tide, a drone equipped with a multispectral sensor and a lidar sensor is used to conduct a low-altitude, water-surface scan of the target planting area to obtain information on the type of tidal flat substrate and the elevation of the tidal flat; and the tidal flat substrate type information is mapped to the Manning roughness coefficient; at high tide, an unmanned surface vessel equipped with an acoustic Doppler current profiler and a multi-parameter water quality probe is used to cruise on the water surface to obtain the water depth, three-dimensional current velocity, salinity, and turbidity of the target planting area.

3. The intelligent mangrove planting method based on unmanned aerial vehicle (UAV) swarms according to claim 2, characterized in that, In step S20, the specific method for generating a mangrove planting navigation map with suitable planting zones and recommended tree species based on the environmental data through hydrodynamic simulation and habitat suitability assessment is as follows: S201, input the tidal flat elevation information, Manning roughness coefficient, water depth, three-dimensional flow velocity, salinity, turbidity and tidal period information into the PINN model, use the two-dimensional shallow water equation as the core loss function of the PINN model for joint training, and inversely retrieve the full-domain continuous hourly water surface elevation field and flow velocity field of the target planting area. S202 calculates the effective water depth based on the water surface elevation field and tidal flat elevation information, and generates a flooding duration map; calculates the bottom shear stress based on the velocity field and Manning roughness coefficient, and generates a water flow scour map; generates a salinity zoning map by performing Kriging spatial interpolation based on unmanned vessel navigation data; and generates a bottom sediment turbidity stability map by inverting the spatiotemporal evolution characteristics of turbidity. S203. The flood duration map, water flow scour map, salinity zoning map and bottom sediment turbidity stability map are used as feature layers and input into the multi-criteria evaluation model. The multi-criteria evaluation model performs pixel-by-pixel spatial algebra operations on the feature layers through a weighted linear combination method, calculates and outputs the comprehensive suitability score of each pixel, and generates a suitability comprehensive score map with continuous score surface based on the comprehensive suitability score of each pixel. S204 divides the suitability comprehensive score map into different levels of partition maps according to the score threshold, and superimposes the salinity tolerance threshold of different mangrove species to match recommended tree species for each partition, and finally generates a smart mangrove planting navigation map.

4. The intelligent mangrove planting method based on unmanned aerial vehicle (UAV) swarms according to claim 3, characterized in that, In step S30, the specific method for assigning tasks and coordinating path planning for the drone swarm performing planting tasks based on the mangrove planting navigation map, and generating the flight path and deployment sequence of each drone, is as follows: S301, extract the coordinates of suitable planting areas from the mangrove planting navigation map as the target point set, and construct the take-off and landing base station, supply point and target point set as a node network for UAV swarm path planning; S302, based on a node network, constructs a multi-UAV path planning model with the goal of minimizing the total flight cost of a UAV swarm. The total flight cost includes flight distance, flight energy consumption, operation time, and environmental risk cost. The constraints of the multi-UAV path planning model include the maximum payload of a single UAV, battery life, maximum range, unique target point allocation, take-off, landing and return, tidal operation time window, safe distance between UAVs, and avoidance of no-fly zones or obstacles. S303 employs an ant colony optimization algorithm to solve the multi-UAV path planning model. By simulating the path search and pheromone update mechanism of artificial ants, it globally optimizes the task allocation scheme that minimizes the total flight cost while satisfying all constraints. Finally, it outputs the target point access sequence, optimal flight path, and deployment sequence for each UAV.

5. The intelligent mangrove planting method based on unmanned aerial vehicle (UAV) swarms according to claim 4, characterized in that, In step S40, the specific method for predicting and dynamically compensating for ballistic deviation based on environmental disturbances during the delivery process is as follows: The drone's current position, deployment altitude, flight speed, attitude angle, and ambient wind speed and direction are acquired in real time. Combined with seedling mass, windward area, air resistance coefficient, and initial angle of the deployment device, the two-dimensional ballistic deviation during the seedling's descent is calculated using an aerodynamic model. The release compensation parameters are dynamically generated based on the offset. These parameters include the trigger advance for controlling early release, the pitch angle compensation for the release channel to compensate for longitudinal offset, the yaw angle compensation for the release channel to compensate for lateral offset, and the flight speed correction at the moment of release.

6. A method for planting mangroves based on a drone swarm according to claim 5, characterized in that, In step S40, the specific method for feedback correction based on visual recognition landing point deviation is as follows: at the moment the seedling enters the water or mud, the UAV-borne visual recognition system captures and identifies the characteristic water splashes or mud splashes stirred up, calculates the geographical coordinates of the actual physical landing point based on the principle of binocular stereo vision, and calculates the landing point deviation value between the actual landing point and the predetermined target point. If the landing point deviation value exceeds the set threshold, the landing point deviation value is used as the feedback correction amount to dynamically update the subsequent seedling deployment compensation parameters.

7. A method for planting mangroves based on a drone swarm according to claim 6, characterized in that, In step S50, the specific method for using a drone swarm to periodically monitor the planted seedlings, assess the seedling survival rate and growth status based on the monitoring data, and generate a replanting plan based on the seedling survival rate and growth status is as follows: S501 utilizes a swarm of drones equipped with multispectral cameras and lidar to periodically patrol and monitor the target planting area, simultaneously acquiring multispectral images and lidar point cloud data of the target planting area. S502 calculates the normalized vegetation index and red edge index based on multispectral images, and uses the inversion of leaf chlorophyll content as a spectral physiological characteristic to assess the physiological health of seedlings; at the same time, it generates a canopy height model based on lidar point cloud to extract plant height and canopy width as a three-dimensional structural feature to assess seedling growth. By integrating spectral physiological features and three-dimensional structural features, and comparing them with a preset growth time series baseline, the seedling survival rate and growth status assessment value of each spatial unit are calculated. S503 involves gridding the monitoring area, calculating the average survival rate and growth rate of each grid, and selecting grids with substandard average survival rate and growth rate as potential replanting areas. Spatial neighborhood connectivity analysis is then performed on the potential replanting areas to filter out isolated dead grids and determine the final replanting areas. Based on the survival rate gap, growth rate, and grid area, the replanting priority and required replanting quantity for each replanting area are calculated.

8. A mangrove intelligent planting system based on a drone swarm, applied to the mangrove intelligent planting method based on a drone swarm as described in any one of claims 1-7, characterized in that, include: The environmental data acquisition module is used to drive drones and unmanned boats to acquire environmental data of the target planting area at different tidal times. The intelligent planning module is used to allocate tasks and plan collaborative paths for a fleet of drones performing planting tasks based on the mangrove planting navigation map, generating the flight paths and deployment sequence of each drone; and it is also used to allocate tasks and plan collaborative paths for a fleet of drones performing planting tasks based on the mangrove planting navigation map, generating the flight paths and deployment sequence of each drone. The precision delivery module is used to drive the drone swarm to perform delivery operations according to the plan. During the delivery process, it performs ballistic deviation prediction and dynamic compensation based on environmental disturbances, and provides feedback correction based on visual recognition of landing point deviation. The monitoring, analysis and decision-making module is used to drive a swarm of drones to periodically monitor the planted seedlings, evaluate the seedling survival rate and growth status based on the monitoring data, and generate a replanting plan based on the seedling survival rate and growth status.

9. An electronic device, characterized in that, Includes at least one processor and at least one memory, wherein, The memory stores computer-readable instructions; The computer-readable instructions are executed by one or more processors, causing the electronic device to implement the intelligent mangrove planting method based on a drone swarm as described in any one of claims 1-7.

10. A storage medium having computer-readable instructions stored thereon, characterized in that, The computer-readable instructions are executed by one or more processors to implement the intelligent mangrove planting method based on a drone swarm as described in any one of claims 1-7.

Images