A 1UGV-nUAV cooperative low-altitude remote sensing motion planning system and operation method suitable for an energy-limited environment

Abstract

This invention discloses a 1UGV-nUAV collaborative low-altitude remote sensing motion planning system and operation method suitable for energy-constrained environments, belonging to the field of agricultural intelligent equipment and multi-robot collaborative control technology. The system includes UGVs, a drone swarm, a charging station module, a control and communication module, and a path planning and task allocation module. The method includes: generating multiple candidate charging stations based on target points in the operation area, further obtaining the optimal combination of charging stations; introducing gravity points to optimize the charging station locations and obtain the shortest path; traversing each optimized charging station according to the shortest path; and generating a multi-UAV collaborative multi-hop trajectory covering the corresponding target points based on a path search algorithm and a drone energy consumption model. This invention significantly improves energy utilization efficiency and operational continuity, enabling multi-target low-altitude remote sensing and path optimization tasks to be completed in unstructured outdoor environments.

Description

Technical Field

[0001] This invention relates to the field of agricultural intelligent equipment and multi-robot collaborative control technology, specifically to a 1UGV-nUAV collaborative low-altitude remote sensing motion planning system and operation method suitable for energy-constrained environments. Background Technology

[0002] With the rapid development of smart agriculture and multi-robot systems, agricultural production is gradually shifting towards automation, precision, and collaboration. Especially in the large-scale seed production and monitoring of hybrid rice, corn, and other high-value crops, the complex field environment, dispersed operation areas, and limited energy supply make it difficult for traditional single-machine operation modes to meet the needs of efficient and continuous field data collection and crop management.

[0003] Existing agricultural remote sensing and monitoring equipment largely relies on single unmanned aerial vehicles (UAVs) to perform tasks. However, UAVs are limited by battery capacity and endurance, often requiring them to frequently return to fixed charging points, leading to mission interruptions, incomplete coverage, and low energy efficiency. Furthermore, the layout of fixed charging stations lacks flexibility and cannot be dynamically optimized based on mission requirements, resulting in reduced overall system operating efficiency.

[0004] In unstructured field environments, while unmanned ground vehicles (UGVs) possess strong payload capacity and endurance, their limited visual coverage makes it difficult to efficiently complete precise positioning, data acquisition, and multi-target collaborative tasks in complex terrain. Most existing studies employ centralized path planning methods, which are computationally intensive and difficult to extend to multi-vehicle collaborative scenarios, failing to achieve energy sharing and collaborative planning between UAVs and ground platforms.

[0005] Currently, there is a lack of an intelligent system capable of enabling UGVs to operate collaboratively with multiple UAVs under energy-constrained conditions, dynamically deploy charging nodes, and simultaneously optimize flight and travel paths. Especially in field environments, a mature technological system has yet to be established that balances energy consumption constraints, low-altitude remote sensing coverage, and motion planning.

[0006] Therefore, there is an urgent need to propose an energy-constrained low-altitude remote sensing motion planning system and method based on the 1UGV-nUAV collaborative structure, so as to realize the autonomous perception, path optimization and energy collaborative replenishment of multiple robots in the field environment, thereby improving the continuity and intelligence level of agricultural remote sensing monitoring and operation. Summary of the Invention

[0007] To address the existing technical problems, this invention provides a 1UGV-nUAV collaborative low-altitude remote sensing motion planning system and operation method suitable for energy-constrained environments. The system has a reasonable structure and an efficient collaborative mechanism, enabling it to complete multi-target low-altitude remote sensing and path optimization tasks in unstructured outdoor environments.

[0008] In a first aspect, this invention proposes a 1UGV-nUAV collaborative low-altitude remote sensing operation method suitable for energy-constrained environments, the method comprising: S1. Generate multiple candidate charging stations based on all target points in the work area; and establish a UAV energy consumption model. S2, model and solve for all candidate charging stations and target points to obtain the combination of charging stations that covers all target points and requires the fewest number of candidate charging stations; S3, by introducing a gravity point, optimizes the position of all charging stations within the charging station combination to obtain the optimized charging stations; S4, based on all optimized charging stations, calculates the shortest path by combining the symmetric traveling salesman algorithm and the genetic algorithm; S5: Traverse each optimized charging station in order according to the shortest path. For each optimized charging station, generate a multi-UAV collaborative multi-hop trajectory covering all target points of the current charging station based on the path search algorithm and UAV energy consumption model. S6, the UGV travels sequentially to the optimized charging station according to the shortest path, and the UAV executes the work task sequentially according to the multi-UAV collaborative multi-hop path corresponding to the current optimized charging station.

[0009] Furthermore, S2 specifically involves: first, modeling the coverage relationship between all candidate charging stations and target points as a bipartite graph; second, based on the bipartite graph, obtaining the combination of candidate charging stations that covers all target points and requires the fewest number of candidate charging stations by solving the minimum hit set problem.

[0010] Furthermore, S3 specifically refers to: S301, using target points as units, obtain the target point coverage circle corresponding to each target point; S302, for each charging station, first connect the charging station and the gravitational point, and combine the maximum flight distance constraint of the UAV to solve for the junction point of the connecting line and the circle covered by each target point; determine the point closest to the gravitational point from all junction points and replace the current charging station; S303, repeat S302 until the positions of all charging stations no longer change, at which point the charging stations are the optimized charging stations.

[0011] Furthermore, the gravity point is the average center of all target points, the median center of all target points, or the starting point of the UGV.

[0012] Furthermore, in S4, the genetic algorithm includes two parts: tournament selection and elite retention strategy.

[0013] Furthermore, in S5, the multi-UAV cooperative multi-hop trajectory corresponding to each optimized charging station is specifically as follows: S501, Obtain the target point corresponding to the current optimized charging station; S502, combining the target point set and the UAV energy consumption model, constructs a multi-machine multi-hop energy constraint search graph with optimized charging stations and target points as network nodes; S503, based on the multi-machine multi-hop energy constraint search graph, uses the A* search algorithm to generate a multi-hop trajectory for the current UAV; the multi-hop trajectory is a path that starts from the optimized charging station, skips to several target points, and returns to the optimized charging station before the power is exhausted; S504, repeat S503 until all target points corresponding to the current optimized charging station are covered and the total number of UAVs does not exceed the available limit, to obtain the number of UAVs required for the current optimized charging station and the corresponding multi-hop trajectory; S505 maps the results of all the multi-hop trajectories in S504 to obtain the complete multi-UAV cooperative multi-hop trajectory under the current optimized charging station.

[0014] Furthermore, before executing S1-S6, multiple remote sensing images covering the entire work area need to be collected using a UAV cluster; based on the remote sensing images, high-precision DSM and TDOM of the work area are obtained through point cloud modeling.

[0015] Secondly, this invention proposes a 1UGV-nUAV collaborative low-altitude remote sensing motion planning system suitable for energy-constrained environments, for implementing the aforementioned 1UGV-nUAV collaborative low-altitude remote sensing operation method suitable for energy-constrained environments.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. By using UGVs as mobile charging hubs, the problem of insufficient battery life for drones in energy-constrained field environments is completely solved, supporting large-scale, long-term continuous operations.

[0017] 2. The adopted ECS-LRSMP hierarchical planning framework decomposes the complex multi-agent cooperative problem into sub-problems such as site selection, routing, and path planning. High-quality solutions are obtained through methods such as gravity optimization, genetic algorithms, and A* search, which significantly reduces the total energy consumption and operation time of the system while ensuring full coverage.

[0018] 3. It enables fully automated, high-precision, safe and reliable take-off, landing and charging of drones on a mobile platform without human intervention, thereby improving the system's autonomy and operational efficiency.

[0019] 4. The entire system has high scalability and adaptability, and can be optimized and adjusted according to different field environments, target point distributions and UAV performance parameters, providing a basis for optimization in practical applications.

[0020] In summary, by introducing dynamic charging nodes, decentralized planning, and a multi-algorithm fusion optimization mechanism, this invention achieves globally optimal path planning and adaptive scheduling for multi-machine collaboration in energy-constrained environments. This system can not only be used for impurity detection and removal in hybrid rice seed production fields, but can also be extended to various field scenarios such as precision agriculture, disaster patrol, forestry monitoring, and energy inspection. Attached Figure Description

[0021] Figure 1 This is a diagram showing the overall framework of the 1UGV-nUAV collaborative operation system proposed in this invention. Figure 2 This is a detailed flowchart of the 1UGV-nUAV collaborative operation method proposed in this invention; Figure 3 A diagram illustrating the overall framework of the path planning and task allocation module; Figure 4 This is a schematic diagram of experimental results according to a specific embodiment of the present invention; Figure 5 This is a diagram illustrating the overall framework of the UAV sub-path planning module. Figure 6 This is a study on the influence of the number of target points and farmland area on experimental results in a specific embodiment of the present invention; Figure 7 This document describes the experimental parameter settings and results for a specific embodiment of the present invention. Detailed Implementation

[0022] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below. Technical features in the various embodiments of the present invention can be combined accordingly without mutual conflict.

[0023] This invention provides a 1UGV-nUAV collaborative low-altitude remote sensing motion planning system suitable for energy-constrained environments, such as... Figure 1 As shown, the system specifically includes the following components: Ground-based unmanned vehicles (UGVs): Serving as ground mobile base stations and charging platforms, UGVs integrate a liftable charging platform, power management unit, steady-state controller, wireless communication device, and RTK-GNSS positioning module. The liftable charging platform includes a support chassis, a lifting bracket, a steady-state controller, and a strain detection device. The lifting bracket raises the charging tray to a set height (e.g., 0.4 meters) during UAV takeoff and landing to provide a safe rotor clearance. The strain detection device accurately senses the contact force during UAV landing (e.g., a set threshold of 10 Newtons) and triggers automatic power-off and charging process switching to ensure safe docking.

[0024] Unmanned Aerial Vehicle (nUAV) swarms consist of several drones equipped with high-precision cameras (such as the Zenmuse P1) and RTK-GNSS modules. They can achieve fully autonomous, high-precision (within ±0.01 meters) take-off and landing on the UGV's charging platform, and perform low-altitude remote sensing data acquisition tasks, such as monitoring and data removal, by controlling the UGV to exchange data and energy.

[0025] Charging Station Module: This module is integrated with the UGV design. The UGV itself is a mobile charging station. During mission execution, it dynamically docks at designated locations based on global planning results, forming temporary charging nodes to provide relay charging and data transfer services for the drone swarm, thereby achieving a dynamic and flexible charging network layout.

[0026] Control and Communication Module: Responsible for low-latency bidirectional communication between UGVs and UAVs within the system, enabling real-time status synchronization, task instruction distribution, and coordinated control.

[0027] Path planning and task allocation module: This is the core algorithm module of the system, built on the Energy Constrained and Charging Station Low-Altitude Remote Sensing Motion Planning (ECS-LRSMP) framework, used to generate globally optimal motion trajectories. This module contains a series of cooperating sub-modules that work together to ensure that the total task cost (such as time and distance) of the system is minimized while satisfying the energy constraints of the UAV.

[0028] The overall process of the path planning and task allocation module is as follows: Figure 3 As shown, it mainly includes the following sub-modules that run in sequence: (1) Voronoi mosaic generation module: Based on the distribution of predefined field target points to be visited (such as plants to be inspected), a set of candidate charging stations (i.e., Voronoi vertices) is generated. Its core principle is to ensure that each target point is within the energy coverage of a certain charging station (i.e., the UAV can access the target point from the station and return safely).

[0029] (2) Hitting Set Optimization Module: The coverage relationship between charging stations and target points is modeled as a bipartite graph. The "minimum hitting set" problem is solved using greedy algorithms, thereby obtaining the combination of charging stations that covers the entire range and has the fewest charging nodes from all Voronoi vertices. This step selects the group of charging nodes with the fewest number of nodes from all candidate charging stations, while ensuring that all target points are covered, thus reducing the number of stations that the UGV needs to stop at.

[0030] (3) Gravity Optimization Module: In order to further shorten the movement path of ground unmanned vehicles (UGVs), this module introduces the concept of gravity points (such as the average center XC, median center MC, or UGV path starting point HC of all target points). The Gravity Optimization Algorithm (GOA) is used as an iterative algorithm to adjust the position of the candidate charging nodes selected in the previous step, so that they can gather towards the gravity center while ensuring coverage.

[0031] In the specific iterative fine-tuning process, the algorithm first calculates the line segments connecting each existing charging node to the set gravitational point. Combining this with the maximum flight distance constraint of the unmanned aerial vehicle (UAV), it uses geometric equations to find the junction point between this line and the preset target point coverage circle. Subsequently, the algorithm determines the point closest to the gravitational point from the generated candidate junction points and uses it as the new node to replace the original charging node. This iterative process of calculating junction points and performing replacements is repeated until all original charging nodes are replaced by new junction points that satisfy the energy coverage constraint and are closest to the gravitational point. Finally, the updated set of charging nodes is output. This optimization effectively reduces the two-dimensional spatial distance between nodes, thereby significantly reducing the total global travel distance of the ground-based unmanned vehicle.

[0032] (4) Symmetric Traveling Salesman Problem (STSP) Solving Module: Taking the optimized charging nodes as cities, this module constructs a path planning model based on the symmetric characteristic that the round-trip distance between nodes for unmanned ground vehicles (UGVs) is equal. This module uses an improved genetic algorithm (GA) to plan the shortest Hamiltonian cycle for the UGV to visit all charging nodes and finally return to the starting point.

[0033] In the specific improvement and solution process, this genetic algorithm combines Tournament Selection and Elite Selection. That is, while repeatedly selecting better individuals through random subsets to reproduce the next generation, the best individual in the current population is directly retained to the next generation without any modification, thereby effectively preventing the loss of the optimal solution in the search process.

[0034] Furthermore, in terms of genetic operators, the algorithm incorporates multiple operations to expand the path search space and generate a new generation of population. The mutation operators are improved to employ multiple techniques such as flipping, swapping, and sliding, while the crossover operators combine ordered crossover, cyclic crossover, and sequence-based crossover strategies for global optimization.

[0035] (5) UAV Sub-path Planning Module: The workflow of this sub-module is as follows: Figure 5 As shown, after determining the charging nodes (i.e., UGV stations), for each charging node, the system first obtains the set of target points to be covered by the UGV station (P={P1, P2, ..., Pn}). Subsequently, the system constructs an UAV energy consumption model by combining the energy consumption and heading characteristics of the UAV, and on this basis, constructs a multi-machine, multi-hop energy constraint search graph with UGV stations and various target points (P1 to Pn) as network nodes.

[0036] After the search graph is constructed, this module iteratively uses the A* search algorithm to explore multi-hop feasible sub-paths for the n UAVs assigned to the station. In each iteration, the A* algorithm generates an energy-feasible multi-hop trajectory for each UAV, starting from the station and visiting several assigned target nodes in an ordered, hop-based manner (e.g., generating a work path with 3 or 4 hops), under the constraint of the maximum energy limit of a single UAV, and finally safely returning to the original station before the power is exhausted. For each planned sub-path, the system performs target coverage and UAV number checks until it ensures that all target points are covered by the planned path and the total number of available UAVs is not exceeded. Finally, the module maps the results of the multi-hop sub-paths generated independently by multiple UAVs and synthesizes them into a complete station trajectory scheme for this UGV charging node, thereby outputting the final global n-UAV cooperative multi-hop flight path trajectory.

[0037] Combining the above sub-modules and Figure 2 The motion planning process of the ECS-LRSMP framework shown below, specifically the execution logic of the path planning and task allocation module, is as follows: After the framework starts, it first inputs the environment configuration parameters (cfgParams), task scenario problem parameters (problem_params), UAV performance data (uav_data), and ground unmanned vehicle performance data (ugv_data), and initializes output variables such as data solution (data_sol) for storing evaluation index values, path solution (path_sol) for recording access sequences and routes, and trajectory visualization charts (figures).

[0038] Then, in Phase 1, the initial set of candidate charging sites (V1), drone operation waypoints and coverage information (wp_c), and a Voronoi visualization chart (figV) are generated and output by calculating Voronoi coverage.

[0039] Next, we move to Phase 2, which solves the set coverage problem (i.e., hit set optimization), outputs the filtered minimum set coverage solution (SolL), a coverage relationship table recording the site and target assignments (scp_table), a simplified charging site set (V2), and a coverage optimization distribution map (figGr), and checks whether the system starting point is included in the simplified site set V2.

[0040] After completing the above initialization and filtering, the system will determine whether a gravity parameter (Gp) is set to control the algorithm to converge towards the center of gravity. If the gravity parameter Gp is ​​set, the system will enter the Gravity Gull optimization algorithm branch, calculate the average center (XC) of the target group, the starting point (HC) or median center (MC) of the ground unmanned vehicle (UGV) path, and generate the spatially fine-tuned horizontal and vertical coordinates (v_x, v_y) of the new charging node. Then, using this optimized node as input, the system will generate the global driving path of the UGV by solving the symmetric traveling salesman problem (TSP-GA) based on a genetic algorithm, and generate the multi-hop operation trajectory of the unmanned aerial vehicle (UAV) based on a search method. The system will construct and save a collaborative solution containing the total distance, total duration and stopping points. If the gravity parameter Gp is ​​not set, the system will execute the gravity-free path branch, directly generate the conventional UGV and UAV paths and save the initial solution.

[0041] Finally, the system integrates the above calculation results, draws and synthesizes the final multi-aircraft cooperative path including waypoints, starting point, calculation vertex, travel route and energy coverage circle, saves all chart files, and returns the final data solution (data_sol), path solution (path_sol) and chart set (figures) to the system, so as to jointly ensure that the total task cost of the system (such as time and distance) is minimized while satisfying the UAV's energy constraints.

[0042] Based on the above system, this invention proposes a 1UGV-nUAV collaborative operation method, the overall process of which is as follows: Figure 2 As shown, the process is divided into four steps: data acquisition and scenario modeling, global path and task allocation, collaborative execution and dynamic charging, and task termination and recycling. The steps include: Step S1: Data Acquisition and Scene Modeling The UAV swarm takes off from the system-defined global starting point (the same initial docking position as the ground-based unmanned vehicles, UGVs). Following pre-defined manually defined survey area boundaries and flight paths planned by the flight software, the swarm navigates to the mission area (e.g., an orchard or hybrid rice field) and performs low-altitude oblique photogrammetry, acquiring multi-angle, high-overlap remote sensing images and simultaneously recording Position and Orientation System (POS) data. After completing the corresponding survey area image acquisition task, the UAVs return directly to the initial global takeoff point for landing and recovery.

[0043] Subsequently, after manually removing and correcting defective images, the collected data is used to generate a high-density 3D point cloud through aerial triangulation, multi-baseline feature matching, and bundle adjustment. This process is then used to create a high-precision digital surface model (DSM) and a true digital orthophoto map (TDOM), providing a 3D geographic information base for subsequent system navigation and motion planning.

[0044] Step S2: Global Path and Task Allocation The control center loads DSM and TDOM data, and the path planning and task allocation module starts the ECS-LRSMP framework for offline computation. S201, based on the target point distribution, run Voronoi mosaicking and hit set optimization to determine a set of minimized charging sites.

[0045] S202, invoke the gravity optimization module to adjust the location of the charging station to shorten the expected path of the UGV.

[0046] S203, run the genetic algorithm to find the optimal TSP path for UGV to access all optimized sites.

[0047] S204, for each site, runs the A* search algorithm to plan an energy-feasible subpath covering nearby target points for each affiliated drone.

[0048] Step S3: Cooperative Execution and Dynamic Charging S301, the UGV automatically travels along the planned TSP path to the first charging station, unfolds the lifting platform to the ready height, and broadcasts a ready signal through the control and communication module.

[0049] S302, the drone swarm at this site performs tasks in sequence: the drones land smoothly on the UGV platform under the guidance of the PID controller, and the charging station module automatically connects to charge after the strain sensor confirms contact; after charging is completed, the UGV raises the platform, and the drones take off vertically to a safe altitude with controlled acceleration, and then switch to the sub-path planned by the A* algorithm to fly to the target point to perform tasks such as remote sensing monitoring.

[0050] The S303 drone monitors its battery level in real time during operation. When the battery level drops below a set threshold, it immediately interrupts the current sub-path and returns to the charging station where the UGV is currently located or is about to arrive to recharge.

[0051] After completing the drone take-off and landing support at the current station, the S304 UGV drives to the next planned station, and so on, forming a collaborative paradigm of "mobile charging - distributed operation".

[0052] Step S4: Task completion and recycling Once all target points have been visited and the task is completed, the swarm of drones returns to the last charging station. The UGVs can either return to their starting point along the shortest path or enter a standby state.

[0053] To verify the effectiveness of the method of the present invention, further experiments were designed. The specific experimental scenarios, experimental parameter settings, and experimental results are as follows: 1. Experimental Scenario and Platform Setup like Figure 7 As shown in (A), the field verification of this system was conducted in hybrid rice seed production fields located in Haining City and Huzhou City, Zhejiang Province. In the system hardware platform, the ground unmanned vehicle (UGV) adopted the IR-Robotic platform (3.64×2.40×2.52 cubic meters); the unmanned aerial vehicle (UAV) adopted the DJI Matrice 300 RTK, equipped with a Zenmuse P1 visible light camera, a Pixhawk 4 autopilot, and an RTK-GNSS positioning module. The simulation and planning of the core algorithm were performed on a computer equipped with an Intel Core i9-10980XE CPU, 62.5 GB of memory, and an NVIDIA RTX A6000 GPU. The experiment divided typical fields of various sizes for scene testing, such as a small verification map with an area of ​​2766 square meters (approximately 4.15 acres) containing 33 weed removal target points, and a large verification map with an area of ​​8456.84 square meters (approximately 12.69 acres) containing 73 target points.

[0054] 2. Experimental Parameter Settings To comprehensively evaluate the performance of the planning system, the core control parameters set include: the number of target points N, the maximum flight distance of the UAV R (used to define energy constraints), and the number of target point clusters. (Used to characterize the spatial distribution of the target). During algorithm performance verification, the number of target points N is often kept constant, and the energy constraint R and spatial distribution characteristics are systematically adjusted. To evaluate the system's adaptability. The maximum flight distance R of the drone was set with multiple gradient values, such as 3km, 6km, and 9km, and the number of clusters was also considered. Multiple gradient values ​​were also set in the range of 3 to 15.

[0055] 3. Experimental Results and Comparative Analysis (1) Overall system performance: For the aforementioned small and large test farmlands, after adopting the ECS-LRSMP collaborative path optimization strategy of this invention, the total task distance was reduced by 15% and 20% respectively compared with the unoptimized initial planning scheme. In the field operation, the success rate of the UAV in a single dynamic landing and charging on the UGV platform was as high as 96% (24 / 25 successful attempts), and the average charging docking position deviation was only 0.1 meters; the actual total operation time was 31.2 minutes, which was significantly reduced by 27% compared with the conventional pure UGV traversal scheme, and no mission failures occurred due to energy depletion; the error between the actual running trajectory and the simulation plan was only 4.7% (simulated total distance 1847 meters, actual measured distance 1938 meters), proving that the system has extremely high planning accuracy and practical feasibility.

[0056] (2) Baseline comparison and ablation experiments: The method of this invention was compared with representative baseline methods such as the traditional fixed charging station strategy (FCURP-MRS) and the pure TSP-GA strategy without gravity optimization. The experimental results are as follows: Figure 4 and Figure 7 As shown in (B), the 1UGV-nUAV algorithm of the present invention reduces the total system task distance by an average of 19.7%, the total driving distance of UGVs by 22.4%, and the total flight distance of UAVs by 15.1%.

[0057] Further ablation experiments clarified the independent contributions of each key module: if the Gravity Optimization (GOA) module is removed from the system, the actual travel distance of the UGV will increase by 18%, and the total mission distance will increase by 14.3%; removing the Hitting-Set screening will lead to an increase of 17.8% in the redundant travel distance of the UGV; and the absence of the Voronoi initial coverage step will significantly reduce the probability of the algorithm finding a feasible solution by 23%.

[0058] (3) Parameter sensitivity analysis: such as Figure 7As shown in (C), the system exhibits stable positive correlation control characteristics as the experimental parameters change. When the maximum flight distance R of the UAV increases by 300% (e.g., from 3km to 9km), the total flight distance of the UAV increases by 35%, the travel distance of the UGV increases by 15%, and the total mission distance of the system increases by 25%; when the clustering density of field targets increases (i.e., the number of clusters), the system's total mission distance increases by 25%. When the number of field target points N is reduced from 15 to 5, the UGV travel distance is reduced by 28%, and the total task distance is reduced by 32%; when the number of field target points N is increased from 10 to 30, the total task distance is increased by 45%.

[0059] like Figure 6 As shown in the figure, the box plot and line graph intuitively reveal the impact of the number of target points on task completion time, and the impact of farmland area on the total energy consumption of the system. The results clearly show that as the number of target points increases or the farmland area expands, both task completion time and total energy consumption exhibit a stable and significant positive correlation growth trend, further verifying the adaptive capability and reliable resource consumption prediction of the system of this invention in dealing with agricultural tasks of different scales and complexities.

[0060] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the invention. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention. Therefore, all technical solutions obtained through equivalent substitution or transformation fall within the protection scope of the present invention.

Claims

1. A method for 1UGV-nUAV collaborative low-altitude remote sensing operations suitable for energy-constrained environments, characterized in that, include: S1, generate multiple candidate charging stations based on all target points in the work area; And establish a UAV energy consumption model; S2, model and solve for all candidate charging stations and target points to obtain the combination of charging stations that covers all target points and requires the fewest number of candidate charging stations; S3, by introducing a gravity point, optimizes the position of all charging stations within the charging station combination to obtain the optimized charging stations; S4, based on all optimized charging stations, calculates the shortest path by combining the symmetric traveling salesman algorithm and the genetic algorithm; S5: Traverse each optimized charging station in order according to the shortest path. For each optimized charging station, generate a multi-UAV collaborative multi-hop trajectory covering all target points of the current charging station based on the path search algorithm and UAV energy consumption model. S6, the UGV travels sequentially to the optimized charging station according to the shortest path, and the UAV executes the work task sequentially according to the multi-UAV collaborative multi-hop path corresponding to the current optimized charging station.

2. The method for 1UGV-nUAV collaborative low-altitude remote sensing operations suitable for energy-constrained environments according to claim 1, characterized in that, S2 specifically involves: first, modeling the coverage relationship between all candidate charging stations and target points as a bipartite graph; second, based on the bipartite graph, obtaining the combination of candidate charging stations that covers all target points and requires the fewest number of candidate charging stations by solving the minimum hit set problem.

3. The method for 1UGV-nUAV collaborative low-altitude remote sensing operations suitable for energy-constrained environments according to claim 1, characterized in that, Specifically, S3 is: S301, using target points as units, obtain the target point coverage circle corresponding to each target point; S302, for each charging station, first connect the charging station and the gravitational point, and combine the maximum flight distance constraint of the UAV to solve for the junction point of the connecting line and the circle covered by each target point; determine the point closest to the gravitational point from all junction points and replace the current charging station; S303, repeat S302 until the positions of all charging stations no longer change, at which point the charging stations are the optimized charging stations.

4. The method for 1UGV-nUAV collaborative low-altitude remote sensing operations suitable for energy-constrained environments according to claim 3, characterized in that, The gravity point is the average center of all target points, the median center of all target points, or the starting point of the UGV.

5. A method for 1UGV-nUAV collaborative low-altitude remote sensing operations suitable for energy-constrained environments according to claim 1, characterized in that, In S4, the genetic algorithm includes two parts: tournament selection and elite retention strategy.

6. The method for 1UGV-nUAV collaborative low-altitude remote sensing operations suitable for energy-constrained environments according to claim 1, characterized in that, In S5, the multi-UAV collaborative multi-hop trajectory corresponding to each optimized charging station is specifically as follows: S501, Obtain the target point corresponding to the current optimized charging station; S502, combining the target point set and the UAV energy consumption model, constructs a multi-machine multi-hop energy constraint search graph with optimized charging stations and target points as network nodes; S503, based on the multi-machine multi-hop energy constraint search graph, uses the A* search algorithm to generate a multi-hop trajectory for the current UAV; the multi-hop trajectory is a path that starts from the optimized charging station, skips to several target points, and returns to the optimized charging station before the power is exhausted; S504, repeat S503 until all target points corresponding to the current optimized charging station are covered and the total number of UAVs does not exceed the available limit, to obtain the number of UAVs required for the current optimized charging station and the corresponding multi-hop trajectory; S505 maps the results of all the multi-hop trajectories in S504 to obtain the complete multi-UAV cooperative multi-hop trajectory under the current optimized charging station.

7. A method for 1UGV-nUAV collaborative low-altitude remote sensing operations suitable for energy-constrained environments according to claim 1, characterized in that, Before executing S1-S6, multiple remote sensing images covering the entire work area need to be collected using a UAV cluster; based on the remote sensing images, high-precision DSM and TDOM of the work area are obtained through point cloud modeling.

8. A 1UGV-nUAV collaborative low-altitude remote sensing motion planning system suitable for energy-constrained environments, used to implement the 1UGV-nUAV collaborative low-altitude remote sensing operation method for energy-constrained environments as described in claim 1, characterized in that, include: A UGV, equipped with a liftable charging platform and power management device, serves as a mobile charging station to provide dynamic energy replenishment for the UAV cluster; A UAV cluster containing several UAVs is used to perform target point operation tasks based on generated multi-hop trajectories; The charging station module is used to be mounted on UGVs to provide relay charging and data transfer services for drone swarms; The control and communication module is used to enable low-latency bidirectional communication and command interaction between the UGV and UAV. The path planning and task allocation module is used to generate the shortest path for the UGV and the corresponding multi-hop trajectory for the UAV.

9. A 1UGV-nUAV collaborative low-altitude remote sensing motion planning system suitable for energy-constrained environments according to claim 8, characterized in that, The path planning and task allocation module includes: The Voronoi mosaic generation module is used to generate multiple candidate charging stations based on all target points in the work area; The hit set optimization module is used to model and solve all candidate charging stations and target points to obtain the combination of charging stations that covers all target points and requires the fewest number of candidate charging stations. The gravity optimization module is used to optimize the position of all charging stations in the charging station combination by introducing gravity points, so as to obtain the optimized charging stations. The Symmetric Traveling Salesman Problem Solving Module is used to calculate the shortest path for UGV based on the optimized charging stations, combining the Symmetric Traveling Salesman Algorithm and the Genetic Algorithm. The UAV sub-path planning module is used to establish a UAV energy consumption model. It sequentially traverses each optimized charging station according to the shortest path. For each optimized charging station, based on the path search algorithm and the UAV energy consumption model, it generates a multi-UAV collaborative multi-hop trajectory covering all target points of the current charging station.

Images