Unmanned aerial vehicle control method, device and equipment and storage medium

Abstract

The invention relates to the technical field of unmanned aerial vehicle control, and discloses an unmanned aerial vehicle control method, device and equipment and a storage medium, and the method comprises the steps: obtaining a target flight scene, and carrying out the scene data collection of the target flight scene, so as to obtain a scene data set; performing three-dimensional modeling based on the scene data set to obtain a three-dimensional scene model; generating a preliminary flight path based on the three-dimensional scene model and a preset obstacle avoidance constraint condition; performing paragraph division on the initial flight path to obtain paragraph feature parameters; generating a preliminary control parameter based on the paragraph feature parameter; integrating the preliminary flight path and the preliminary control parameters into a flight control instruction; according to the invention, based on the complete process of scene perception, three-dimensional modeling, path generation, paragraph feature extraction, parameter matching and instruction integration, the flight control instruction adaptive to the actual flight demand is generated, the problem of poor adaptability of the flight path and the control parameter in the prior art is solved, and the flight stability and accuracy of the unmanned aerial vehicle are improved.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) control technology, and in particular to a UAV control method, apparatus, device, and storage medium. Background Technology

[0002] In existing UAV control methods, flight path planning and control parameter generation are mostly independent processes. The parameters are not fully combined with the actual environmental characteristics of the target flight scenario, resulting in poor adaptability between the generated control parameters and the path. This can easily lead to problems such as flight instability and insufficient obstacle avoidance accuracy, and cannot meet the needs of UAVs for precise flight in complex scenarios. Summary of the Invention

[0003] To overcome the shortcomings of existing technologies, the present invention aims to provide a method, device, equipment, and storage medium for controlling unmanned aerial vehicles (UAVs). Based on a complete process of scene perception, 3D modeling, path generation, segment feature extraction, parameter matching, and instruction integration, the present invention generates flight control instructions that are adapted to actual flight requirements, solving the problem of poor adaptability between flight paths and control parameters in existing technologies, and improving the flight stability and accuracy of UAVs.

[0004] The first aspect of the present invention provides a method for controlling an unmanned aerial vehicle (UAV), the method comprising the steps of: acquiring a target flight scene and collecting scene data of the target flight scene to obtain a scene dataset; performing three-dimensional modeling based on the scene dataset to obtain a three-dimensional scene model; generating a preliminary flight path based on the three-dimensional scene model and preset obstacle avoidance constraints; dividing the preliminary flight path into segments to obtain segment feature parameters; generating preliminary control parameters based on the segment feature parameters; and integrating the preliminary flight path and the preliminary control parameters into a flight control command.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the step of acquiring the target flight scene and collecting scene data from the target flight scene to obtain a scene dataset includes: acquiring the target flight scene and determining the start and end points of the flight mission based on the target flight scene to delineate the data collection range; collecting static environmental data based on the data collection range to obtain terrain data and obstacle data; collecting dynamic constraint data based on the data collection range to obtain airspace constraint data and wind speed baseline data; and integrating the terrain data, obstacle data, airspace constraint data, and wind speed baseline data into a scene dataset.

[0006] Optionally, in a second implementation of the first aspect of the present invention, the step of performing three-dimensional modeling based on the scene dataset to obtain a three-dimensional scene model includes: extracting terrain data, obstacle data, spatial constraint data, and wind speed baseline data from the scene dataset; constructing a terrain layer, an obstacle layer, and a spatial constraint layer based on the terrain data, obstacle data, and spatial constraint data, respectively; aligning and superimposing the terrain layer, obstacle layer, and spatial constraint layer according to spatial coordinates to generate a three-dimensional basic model; and using the wind speed baseline data to label wind speed parameters for each spatial region in the three-dimensional basic model to obtain a three-dimensional scene model.

[0007] Optionally, in a third implementation of the first aspect of the present invention, generating a preliminary flight path based on a three-dimensional scene model and preset obstacle avoidance constraints includes: determining the spatial boundary for constructing a path search based on the three-dimensional scene model; generating a feasible path domain based on preset obstacle avoidance constraints and the spatial boundary; and performing a path search within the feasible path domain to generate a preliminary flight path.

[0008] Optionally, in the fourth implementation of the first aspect of the present invention, the step of dividing the preliminary flight path into segments to obtain segment feature parameters includes: dividing the preliminary flight path into segments based on preset segmentation rules to obtain multiple flight segments; identifying the average slope, path curvature, and obstacle spacing of each flight segment to obtain feature parameters of each flight segment; and integrating the feature parameters of each flight segment into segment feature parameters according to the order of the flight segments.

[0009] Optionally, in a fifth implementation of the first aspect of the present invention, the step of generating preliminary control parameters based on paragraph feature parameters includes: mapping paragraph feature parameters to initial control parameters based on a preset mapping rule; obtaining performance threshold parameters of the UAV; and numerically correcting the out-of-limit portion of the initial control parameters based on the performance threshold parameters to obtain preliminary control parameters.

[0010] Optionally, in a sixth implementation of the first aspect of the present invention, the step of integrating the preliminary flight path and preliminary control parameters into flight control commands includes: taking flight segments as units, binding the preliminary control parameters of each segment to the preliminary flight path, calculating the parameter differences between adjacent segments, performing gradient compensation for speed and steering angle for segments with differences exceeding a preset threshold, so as to obtain a path parameter set; and performing command conversion based on the path parameter set to obtain flight control commands.

[0011] A second aspect of the present invention provides a drone control device, comprising: a data acquisition module for acquiring a target flight scene and collecting scene data from the target flight scene to obtain a scene dataset; a model building module for performing three-dimensional modeling based on the scene dataset to obtain a three-dimensional scene model; a path generation module for generating a preliminary flight path based on the three-dimensional scene model and preset obstacle avoidance constraints; a segmentation module for segmenting the preliminary flight path to obtain segment feature parameters; a parameter generation module for generating preliminary control parameters based on the segment feature parameters; and a command generation module for integrating the preliminary flight path and preliminary control parameters into flight control commands.

[0012] A third aspect of the present invention provides a drone control device, the drone control device comprising: a memory and at least one processor, the memory storing instructions; at least one processor calling the instructions in the memory to cause the computer device to execute the various steps of any of the drone control methods described above.

[0013] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the steps of any of the above-described unmanned aerial vehicle (UAV) control methods.

[0014] In the technical solution of this invention, firstly, by acquiring the target flight scene and collecting scene datasets, a comprehensive perception of the flight environment is achieved, providing an environmental foundation for subsequent path planning and parameter generation. Secondly, a three-dimensional scene model is constructed based on the scene dataset, enabling path planning to be carried out in a three-dimensional space that conforms to the actual environment. The preliminary flight path generated in combination with preset obstacle avoidance constraints is more reasonable and safer. Thirdly, the preliminary flight path is divided into segments and segment feature parameters are extracted, achieving a refined decomposition of the path, allowing the generation of control parameters to be specifically adapted to the path characteristics of different segments. Finally, preliminary control parameters are generated based on the segment feature parameters, and the preliminary flight path and preliminary control parameters are integrated into flight control commands, achieving a deep binding between the flight path and control parameters. This invention generates flight control commands adapted to actual flight needs through a complete process of scene perception, three-dimensional modeling, path generation, segment feature extraction, parameter matching, and command integration, solving the problem of poor adaptability between flight paths and control parameters in the prior art, and improving the flight stability and accuracy of UAVs. Attached Figure Description

[0015] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a first flowchart of a drone control method provided in an embodiment of the present invention; Figure 2 This is a second flowchart of the unmanned aerial vehicle (UAV) control method provided in an embodiment of the present invention; Figure 3 This is a third flowchart of the unmanned aerial vehicle (UAV) control method provided in an embodiment of the present invention; Figure 4 This is a fourth flowchart of the unmanned aerial vehicle (UAV) control method provided in an embodiment of the present invention; Figure 5 A fifth flowchart of the unmanned aerial vehicle (UAV) control method provided in an embodiment of the present invention; Figure 6 A sixth flowchart of the unmanned aerial vehicle (UAV) control method provided in an embodiment of the present invention; Figure 7 A seventh flowchart of the unmanned aerial vehicle (UAV) control method provided in an embodiment of the present invention; Figure 8 This is a schematic diagram of the structure of the unmanned aerial vehicle (UAV) control device provided in an embodiment of the present invention; Figure 9 This is a schematic diagram of the structure of a drone control device provided in an embodiment of the present invention. Detailed Implementation

[0016] This invention provides a method, device, equipment, and storage medium for controlling unmanned aerial vehicles (UAVs). Based on a complete process of scene perception, 3D modeling, path generation, segment feature extraction, parameter matching, and command integration, it generates flight control commands that are adapted to actual flight requirements. This solves the problem of poor adaptability between flight paths and control parameters in existing technologies and improves the flight stability and accuracy of UAVs.

[0017] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0018] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the unmanned aerial vehicle (UAV) control method in this invention includes: The drone control method includes the following steps: 101. Obtain the target flight scene and collect scene data from the target flight scene to obtain a scene dataset; In this embodiment, the target flight scenario for the mission to be performed by the UAV is first defined, and the start and end points of the flight are determined to define the collection range. Then, static environmental data and dynamic constraint data within this range are collected. The static environmental data includes terrain data and obstacle data, and the dynamic constraint data includes airspace constraint data and wind speed baseline data. Finally, the various types of collected data are integrated to form a complete scene dataset.

[0019] 102. Perform 3D modeling based on the scene dataset to obtain a 3D scene model; In this embodiment, corresponding terrain layers, obstacle layers, and airspace layers are constructed based on terrain data, obstacle data, and airspace constraint data, respectively. The three-layer model is superimposed to form a three-dimensional basic model. Then, the wind speed baseline data is labeled to each spatial region of the three-dimensional basic model to finally obtain a three-dimensional scene model.

[0020] 103. Generate a preliminary flight path based on the 3D scene model and preset obstacle avoidance constraints; In this embodiment, the spatial boundary of the path search is determined based on the three-dimensional scene model, and then the feasible region of the path is selected by combining the preset obstacle avoidance constraints. Within the feasible region, the path search algorithm is used to plan the path from the starting point to the end point, and a preliminary flight path that meets the safety obstacle avoidance requirements is generated.

[0021] 104. Divide the preliminary flight path into segments to obtain segment characteristic parameters; In this embodiment, the flight path is first divided into segments based on the curvature changes and abrupt slope changes of the initial flight path; then, the average slope, path curvature, and distance between surrounding obstacles are extracted for each flight segment to obtain the characteristic parameters of each flight segment; and finally, the characteristic parameters of each flight segment are integrated to form the segment characteristic parameters.

[0022] 105. Generate preliminary control parameters based on paragraph feature parameters; In this embodiment, a mapping rule is preset between paragraph feature parameters and UAV control parameters (such as flight speed, climb rate, and turn rate). The initial control parameters are obtained by substituting the feature parameters of each paragraph into the rule.

[0023] 106. Integrate the preliminary flight path and preliminary control parameters into flight control commands; In this embodiment, the initial control parameters of each flight segment are precisely bound to the initial flight path to form a path parameter group. Then, the path parameter group is converted into instructions according to the communication protocol supported by the UAV flight control system to generate flight control instructions that can be directly recognized and executed by the flight control system.

[0024] In this embodiment of the invention, firstly, by acquiring the target flight scene and collecting scene datasets, a comprehensive perception of the flight environment is achieved, providing an environmental foundation for subsequent path planning and parameter generation. Secondly, a three-dimensional scene model is constructed based on the scene dataset, enabling path planning to be carried out in a three-dimensional space that conforms to the actual environment. The preliminary flight path generated in combination with preset obstacle avoidance constraints is more reasonable and safer. Thirdly, the preliminary flight path is divided into segments and segment feature parameters are extracted, achieving a refined decomposition of the path, allowing the generation of control parameters to be specifically adapted to the path characteristics of different segments. Finally, preliminary control parameters are generated based on the segment feature parameters, and the preliminary flight path and preliminary control parameters are integrated into flight control commands, achieving a deep binding between the flight path and control parameters. This invention generates flight control commands adapted to actual flight needs through a complete process of scene perception, three-dimensional modeling, path generation, segment feature extraction, parameter matching, and command integration, solving the problem of poor adaptability between flight paths and control parameters in the prior art, and improving the flight stability and accuracy of UAVs.

[0025] Please see Figure 2 Two embodiments of the UAV control method in this invention include step 101, which includes: 201. Obtain the target flight scene and determine the start and end points of the flight mission based on the target flight scene to define the data collection range; In this embodiment, the target flight scene is divided into known fixed scene and unknown new scene; if it is a known fixed scene, the pre-stored basic scene information can be directly called; if it is an unknown new scene, the high-definition camera on the drone can be used to conduct preliminary aerial reconnaissance to obtain the approximate range and outline information of the scene. When determining the start and end points of a flight mission, it is necessary to specify the operational coordinates that the UAV needs to reach, based on the actual mission requirements (such as material transportation, power line inspection, terrain survey, etc.). When defining the data collection area, the "start-end core line expansion method" is adopted. Taking the line connecting the start and end points as the central axis, a rectangular collection area is formed by extending 50-100 meters to both sides. The expansion distance can be dynamically adjusted according to the complexity of the scene. For complex terrain (such as mountains and urban building complexes), the expansion range can be appropriately increased to 100-200 meters, while for simple terrain (such as plains and open farmland), it can be reduced to 30-50 meters. This collection method ensures that the collection range fully covers the key environmental information around the flight path, avoids missing important obstacles or terrain undulation data due to the collection range being too small, and at the same time prevents data redundancy caused by the range being too large, which would increase the pressure of subsequent modeling and calculation.

[0026] 202. Conduct static environmental data collection based on the data collection range to obtain terrain data and obstacle data; In this embodiment, during terrain data acquisition, the UAV is equipped with a lidar sensor and a high-precision GPS module. It flies within the acquisition range according to a preset grid route, with the grid spacing set at 20 meters. The flight altitude is adjusted according to the scene altitude to ensure that the lidar's detection range can completely cover the acquisition area. The lidar sensor calculates the three-dimensional coordinates of each point on the ground by emitting laser pulses and receiving reflected signals. Combined with GPS positioning information, it generates high-density point cloud data. Then, noise points are removed by a point cloud filtering algorithm (such as pass-through filtering) to finally obtain accurate terrain data. Obstacle data collection and terrain data collection are carried out simultaneously. During aerial photography, the UAV uses high-definition cameras and lidar to detect and identify fixed obstacles (such as buildings, trees, towers, bridges, etc.) within the collection range, and records the three-dimensional coordinates, size, shape and location distribution information of the obstacles. For lidar detection blind spots (such as the bottom of obstacles), ground personnel can carry handheld laser rangefinders for supplementary surveys. Static environmental data (terrain, obstacles) is the core foundation for UAV flight path planning. Its accuracy directly affects the safety and rationality of the path, providing reliable data support for the subsequent construction of accurate 3D scene models.

[0027] 203. Dynamically constrain data acquisition is performed based on the data acquisition range to obtain airspace constraint data and wind speed baseline data; In this embodiment, airspace constraint data within the collection range is obtained through the airspace control system of the national airspace management department, including the scope and altitude restrictions of no-fly zones, restricted flight zones, and controlled airspace, as well as airspace usage rights during flight periods. Wind speed baseline data was collected using a miniature meteorological sensor mounted on a drone. The drone flew along a preset route within the collection area, and the miniature meteorological sensor collected wind speed and direction data in real time at different altitudes (each altitude layer was 50 meters apart). During drone flight, airspace constraints directly determine the legality of the flight path, while wind speed changes affect the drone's flight attitude and stability; therefore, dynamic constraint data is a necessary condition for generating safe and feasible flight paths.

[0028] 204. Integrate terrain data, obstacle data, airspace constraint data, and wind speed baseline data into a scene dataset; In this embodiment, terrain data, obstacle data, airspace constraint data, and wind speed baseline data are uniformly converted into the WGS-84 geodetic coordinate system to ensure that the spatial coordinate reference of all data is consistent and to avoid data misalignment caused by differences in coordinate systems. All the converted data are integrated into a scene dataset and stored in the database of the UAV ground control station.

[0029] Please see Figure 3In the three embodiments of the UAV control method of the present invention, step 102 includes: 301. Extract terrain data, obstacle data, airspace constraint data, and wind speed baseline data from the scene dataset.

[0030] 302. Construct terrain layer, obstacle layer, and airspace layer based on terrain data, obstacle data, and airspace constraint data, respectively; In this embodiment, a suitable modeling algorithm is selected based on the characteristics of different environmental factors to ensure the accuracy and practicality of the models at each layer. The terrain layer is constructed using the Triangular Irregular Network (TIN) modeling algorithm: First, the terrain point cloud data is resampled to adjust the point cloud density to 4-6 sampling points per square meter. Then, key terrain feature points such as mountain tops, valleys, and slope toes are identified through feature extraction algorithms. Using these feature points as core nodes, the Delaunay triangulation algorithm is used to construct a continuous terrain triangulation network, transforming the discrete point cloud data into a terrain layer that can accurately reflect the terrain undulation characteristics. The obstacle layer construction includes regular entity modeling and irregular surface modeling. For regular-shaped obstacles such as buildings and towers, corresponding regular entity models such as cuboids and cylinders are constructed based on their three-dimensional coordinates and size parameters. For irregular-shaped obstacles such as trees and mountain slopes, a moving least squares surface reconstruction algorithm is used to construct a surface model that fits the actual shape based on obstacle point cloud data. Finally, the regular entity model and the surface model are integrated to form a complete obstacle layer. The airspace layer is constructed using a spatial region division method: First, based on the range coordinates of no-fly zones, restricted-fly zones, and controlled airspace in the airspace constraint data, the corresponding regional boundaries are delineated in three-dimensional space; then, a differentiated marking strategy is adopted, with no-fly zones marked with red semi-transparent surfaces, restricted-fly zones marked with yellow semi-transparent surfaces, and controlled airspace marked with blue semi-transparent surfaces. At the same time, attribute labels such as altitude restrictions and flight time restrictions are added to each airspace region to form an airspace layer. Different environmental elements have different physical forms and functional attributes. Targeted modeling can maximize the advantages of each algorithm. Each layer of the model accurately matches the actual characteristics of the corresponding environmental elements. For subsequent path planning, it is necessary to clearly distinguish different environmental constraints such as terrain, obstacles, and airspace constraints.

[0031] 303. Align and overlay the terrain layer, obstacle layer, and airspace layer according to spatial coordinates to generate a three-dimensional basic model; In this embodiment, the terrain layer is first loaded into the 3D modeling space as the basic carrier of the entire model; then the obstacle layer is loaded, and the obstacles are precisely placed at the corresponding surface positions of the terrain layer according to their 3D coordinates; finally, the airspace layer is loaded to completely cover the entire modeling space; during the stacking process, a collision detection algorithm is enabled to check whether there are spatial conflicts between the models of each layer, such as the top of the obstacle exceeding the height limit of the corresponding airspace area, or the airspace area boundary not matching the terrain boundary; if a conflict is detected, the process returns to the construction step of the corresponding layer, corrects the model parameters (such as adjusting the obstacle height and correcting the airspace boundary coordinates), and then stacks again until there are no spatial conflicts; finally, the stacked models of each layer are integrated into a whole to generate a 3D basic model containing three core static environmental elements: terrain, obstacles, and airspace constraints.

[0032] 304. Use wind speed baseline data to label wind speed parameters for each spatial region in the 3D basic model to obtain a 3D scene model; In this embodiment, the wind speed baseline data is first preprocessed and categorized by height layers (a standard height layer is set every 50 meters) to form a relationship table of "height layer-wind speed-wind direction". Then, the 3D base model is divided into multiple height layers parallel to the ground at equal height intervals, establishing a correspondence between the model height layers and the wind speed data height layers. The wind speed and direction data for the corresponding height layers are labeled to the corresponding spatial areas of the model. For the transition area between adjacent standard height layers (such as any height position between 50 meters and 100 meters), a linear interpolation algorithm is used to calculate the wind speed parameters: Let the wind speed at the lower standard height layer (50 meters) be v1, and the wind speed at the upper standard height layer (100 meters) be v1. If the wind speed is v2, and the distance between the transition zone height and the lower standard layer is h1, and the distance between the transition zone height and the upper standard layer is h2, then the wind speed in the transition zone is v = v1 × (h2 / (h1 + h2)) + v2 × (h1 / (h1 + h2)). This ensures a smooth transition of wind speed parameters in the transition zone, avoiding abrupt parameter changes. After annotation, a full-area traversal check is performed on the 3D basic model. If there are local areas with missing wind speed data, such as the edges of the data collection range, wind speed data from adjacent valid areas is used for supplementary annotation to ensure no wind speed parameters are missed. Finally, the 3D basic model with completed wind speed parameter annotation is determined as a complete 3D scene model and stored in the model database of the UAV ground control station. Wind speed data has significant height stratification characteristics. First, basic annotation is achieved through height layer matching. Then, interpolation is used to supplement the annotation of non-standard height layers and data missing areas, so that the 3D scene model contains complete wind speed environment information, providing data support for the subsequent generation of flight paths and control parameters adapted to wind speed conditions.

[0033] Please see Figure 4 In the four embodiments of the UAV control method of the present invention, step 103 includes: 401. Determine the spatial boundaries of the path search based on the 3D scene model; In this embodiment, firstly, the three-dimensional coordinates of the starting point and the ending point of the flight mission are extracted from the three-dimensional scene model, and the core area of ​​the path search is locked with the line connecting the starting point and the ending point as the core axis. Subsequently, a safe buffer distance was set based on the terrain, obstacles, and spatial constraints in the 3D scene model: for simple terrains such as plains, the buffer distance was extended by 50 meters to each side based on the core axis; for complex terrains such as mountainous areas and urban building complexes, the buffer distance was extended by 100-150 meters to each side, taking into account the large terrain undulations and dense distribution of obstacles. At the same time, referring to the height restriction information in the airspace layer, the vertical height boundary of the path search is determined, that is, the minimum flight height is not lower than 50 meters above the top of the highest obstacle on the ground, and the maximum flight height does not exceed the maximum allowable flight height marked in the airspace layer; Ultimately, by combining the core area, the horizontal safety buffer distance, and the vertical height boundary, a closed path search space boundary is formed.

[0034] 402. Generate a feasible path region based on preset obstacle avoidance constraints and spatial boundaries; In this embodiment, firstly, the preset obstacle avoidance constraints include hard obstacle avoidance constraints and flexible obstacle avoidance constraints: the hard obstacle avoidance constraints are that the minimum safe distance between the UAV and the obstacle is not less than 30 meters and the flight altitude does not exceed the limit range marked by the airspace layer; the flexible obstacle avoidance constraints are to prioritize avoiding areas with wind speeds greater than level 6. Subsequently, based on the spatial boundary determined in step 401, the regions within this range are filtered point by point in the 3D scene model: regions that violate hard obstacle avoidance constraints are eliminated, and high-wind-speed regions that violate flexible obstacle avoidance constraints are marked as "non-priority selection regions"; finally, the regions that meet all constraints after filtering are integrated into a continuous path feasible region, and the marking information of "non-priority selection regions" is associated with the feasible region to provide a reference for subsequent path search; The feasible region is the direct range of subsequent path search, and its accuracy directly determines the safety and feasibility of the generated path. It is necessary to ensure that all areas within the feasible region meet flight safety requirements through strict constraint filtering. In this embodiment, all dangerous areas are eliminated through preset obstacle avoidance constraints, and "non-priority selection areas" are marked to provide optimization directions for path search.

[0035] 403. Perform a path search within the feasible region to generate a preliminary flight path; In this embodiment, path search employs an improved A... Algorithm; First, set A The heuristic function h(n) of the algorithm: Heuristic function h(n) = α × d(n, goal) + β × e(n); Where d(n,goal) is the straight-line distance from the current node to the destination, e(n) is the flight energy consumption of the area where the current node is located (calculated based on wind speed parameters; the higher the wind speed, the higher the energy consumption), and α and β are weight coefficients, set to 0.6 and 0.4 respectively, to ensure that the algorithm prioritizes searching for paths with shorter paths and lower energy consumption; Subsequently, with the starting point of the flight mission as the search starting point and the ending point as the target point, a node expansion search is performed within the feasible path region: starting from the starting point, multiple adjacent feasible nodes (nodes that meet the obstacle avoidance constraints) are expanded each time. The cost function f(n) = g(n) + h(n) of each node is calculated through a heuristic function (g(n) is the actual path length from the starting point to the current node). The node with the smallest cost function is selected as the next search node, and this process is iterated until the end point is reached. During the search process, high-wind-speed areas marked as "non-priority selection areas" are avoided. If it is not possible to completely avoid them, the path segment with the smallest wind speed is selected to pass through. After the search is completed, an initial path from the starting point to the end point is obtained. The initial path is then smoothed to remove sharp inflection points (by optimizing the path curve using a Bézier curve fitting algorithm), and finally a continuous and smooth preliminary flight path is generated. The preliminary flight path not only meets safety requirements such as obstacle avoidance and airspace constraints, but also features a short path, low energy consumption, and smooth continuity, making it suitable for the flight characteristics of UAVs.

[0036] Please see Figure 5 Five embodiments of the UAV control method in this invention include step 104, which includes: 501. Divide the initial flight path into segments based on preset segmentation rules to obtain multiple flight segments; In this embodiment, firstly, the initial flight path is discretized into path points spaced 0.5 meters apart, and the curvature value of each path point is calculated (curvature calculation uses the three-point method, fitting an arc with the coordinates of the current point and two adjacent path points, and deriving the curvature from the arc radius). The curvature threshold is 0.03 rad / m. When the curvature value of a path point exceeds the threshold, it is determined to be a feature inflection point and automatically marked as a segment node. For continuous path segments without feature inflection points, a fixed length safety threshold is set (400 meters in this embodiment, which can be dynamically adjusted according to the accuracy of the flight mission). If the length of a continuous path segment without inflection points exceeds the threshold, a segment node is forcibly set at the center point of the path segment. Finally, based on all marked segment nodes, the initial flight path is divided into multiple continuous flight segments.

[0037] 502. Identify the average slope, path curvature, and obstacle spacing of each flight segment to obtain the characteristic parameters of each flight segment; In this embodiment, when identifying the average slope, the three-dimensional coordinates of the segment nodes at both ends of each flight segment are first extracted, the elevation difference ΔZ between the two ends of the node is calculated, and then the horizontal distance D of the segment is calculated using the formula for the straight-line distance between two points, so that the average slope α (tanα=ΔZ / D) can be obtained. When performing path curvature identification, the path points of a single flight segment are first fitted into a continuous and smooth curve using a cubic B-spline curve (to reduce the influence of discrete point errors). Then, at least 20 sampling points are uniformly selected on the fitted curve (the number of sampling points can be dynamically adjusted according to the segment length). The curvature value of each sampling point is calculated. Finally, the arithmetic mean of the curvature values ​​of all sampling points is taken as the path curvature parameter of the flight segment. When identifying obstacle spacing, the system iterates through all path points in the current flight segment and calculates the straight-line distances to surrounding obstacles (buildings, trees, towers, etc.), selecting the minimum value as the obstacle spacing for that segment. If the distance from all path points to obstacles is greater than 500 meters, the default obstacle spacing is 500 meters. In this embodiment, the average slope reflects the elevation and descent requirements of the segment, the path curvature reflects the turning requirements, and the obstacle spacing reflects the obstacle avoidance safety requirements. These three parameters directly determine the power output, steering control, and obstacle avoidance strategy of the UAV in that segment, and are the core basis for generating subsequent control parameters.

[0038] 503. Integrate the characteristic parameters of each flight segment into segment characteristic parameters according to the order of the flight segments.

[0039] Please see Figure 6 The six embodiments of the UAV control method in this invention include step 105, which includes: 601. Based on preset mapping rules, map paragraph feature parameters to initial control parameters; In this embodiment, the initial control parameters include flight speed, climb rate, and yaw rate, and their corresponding mapping rules are as follows: Flight speed mapping rule: When the distance between obstacles is large and the path curvature is small, a larger flight speed corresponds to the flight speed; when the distance between obstacles is small or the path curvature is large, a smaller flight speed corresponds to the flight speed. The specific flight speed mapping rules are as follows: When the obstacle spacing is greater than 300m and the path curvature is less than 0.02rad / m, the preset flight speed is 10m / s; when the obstacle spacing is 200-300m and the path curvature is 0.02-0.04rad / m, the preset flight speed is 7m / s; when the obstacle spacing is less than 200m or the path curvature is greater than 0.04rad / m, the preset flight speed is 4m / s. Ascent and descent speed mapping rule: When the average slope is upward and the slope is large, it corresponds to a large ascent speed; when the average slope is downward and the slope is large, it corresponds to a large descent speed. The specific rules for mapping the elevation and descent speeds are as follows: When the average slope increases and is greater than 5°, the preset elevation and descent speed is 1.0 m / s; when the average slope increases and is between 2° and 5°, the preset elevation and descent speed is 0.6 m / s; when the average slope is horizontal (within ±2°), the preset elevation and descent speed is 0 m / s; when the average slope decreases and is between 2° and 5°, the preset elevation and descent speed is -0.6 m / s; when the average slope decreases and is greater than 5°, the preset elevation and descent speed is -1.0 m / s. Steering angular velocity mapping rule: The greater the path curvature, the greater the steering angular velocity; The specific rules for mapping steering angular velocity are as follows: when the path curvature is less than 0.02 rad / m, the steering angular velocity is preset to 0.08 rad / s; when the path curvature is 0.02-0.04 rad / m, the steering angular velocity is preset to 0.15 rad / s; when the path curvature is greater than 0.04 rad / m, the steering angular velocity is preset to 0.25 rad / s. The mapping method is as follows: call the paragraph feature parameter matrix, extract the three feature parameters of average slope, path curvature and obstacle spacing of each paragraph in sequence according to the flight paragraph number, match the extracted parameters with the above preset numerical relationship, and obtain the initial control parameters corresponding to the paragraph after successful matching.

[0040] 602. Obtain the performance threshold parameters of the UAV, and based on the performance threshold parameters, perform numerical correction on the out-of-limit part of the initial control parameters to obtain the preliminary control parameters; In this embodiment, firstly, the performance threshold parameters of the UAV are obtained. These parameters are derived from the UAV's manufacturer's technical specifications and the preset configuration of the ground control station; specifically, they include the maximum flight speed, minimum flight speed, maximum climb rate, minimum climb rate, maximum angular velocity, and minimum angular velocity. Subsequently, each initial control parameter is checked one by one to see if it is within the corresponding performance threshold range: if the initial parameter is within the threshold range, it is directly retained; if it exceeds the threshold, gradient correction is performed to correct the out-of-limit parameter to the corresponding upper or lower threshold limit; finally, all the checked and corrected control parameters are re-bound with the corresponding flight segment number to obtain the preliminary control parameters, which are then stored in the ground control station database. This embodiment is based on the actual performance limits of the UAV. Infeasible parameters are eliminated through threshold verification to avoid damaging the UAV or affecting flight stability.

[0041] Please see Figure 7 The seven embodiments of the UAV control method in this invention include step 106, which includes: 701. Taking flight segments as units, bind the preliminary control parameters of each segment to the preliminary flight path, calculate the parameter difference between adjacent segments, and perform gradient compensation for speed and steering angle for segments with differences exceeding a preset threshold to obtain the path parameter set. In this embodiment, firstly, the preliminary control parameters are bound to the preliminary flight path. Taking the flight segments divided in step 501 as the basic unit, the preliminary control parameters (flight speed, climb rate, and turning angular velocity) of each segment generated in step 602 are associated with the path information such as the three-dimensional coordinates (start point, end point, and critical path point) and segment length of the corresponding numbered segment in the preliminary flight path to form multiple "path parameter units". Secondly, parameter difference calculation and gradient compensation are performed between adjacent segments. First, a parameter smoothing threshold is preset based on the UAV's flight attitude adjustment capability. This threshold is the maximum parameter change range that the UAV can smoothly withstand, ensuring that sudden parameter changes do not cause drastic attitude fluctuations. Then, the differences between the corresponding speed parameters (flight speed, climb rate) and turning angular velocity parameters in two adjacent "path parameter units" are calculated sequentially. If the difference of a certain type of parameter exceeds the preset smoothing threshold, the gradient compensation mechanism is activated: using the parameters of the adjacent preceding segment as a benchmark, the corresponding parameters of the following segment are adjusted according to a preset compensation step size, so that the parameter difference between adjacent segments is reduced to within the threshold range after adjustment. Finally, all "path parameter units" that have completed binding and compensation optimization are integrated in sequence to form a complete path parameter group.

[0042] 702. Perform command conversion based on path parameter group to obtain flight control commands; In this embodiment, the path parameter group is converted into instructions according to the communication protocol supported by the UAV flight control system to generate flight control instructions that can be directly recognized and executed by the flight control system.

[0043] The above describes the UAV control method in the embodiments of the present invention. The following describes the UAV control device in the embodiments of the present invention. Please refer to [link / reference]. Figure 8 One embodiment of the unmanned aerial vehicle (UAV) control device in this invention includes: The data acquisition module 801 is used to acquire the target flight scene and collect scene data from the target flight scene to obtain a scene dataset; The model building module 802 is used to perform 3D modeling based on the scene dataset to obtain a 3D scene model; The path generation module 803 is used to generate a preliminary flight path based on the 3D scene model and preset obstacle avoidance constraints. The segmentation module 804 is used to segment the preliminary flight path to obtain segment feature parameters; The parameter generation module 805 is used to generate preliminary control parameters based on paragraph feature parameters; The instruction generation module 806 is used to integrate the preliminary flight path and preliminary control parameters into flight control instructions; In this embodiment, firstly, the data acquisition module 801 acquires the target flight scene and collects the scene dataset, achieving comprehensive perception of the flight environment and providing an environmental foundation for subsequent path planning and parameter generation. Secondly, the model building module 802 constructs a three-dimensional scene model based on the scene dataset, enabling path planning to be carried out in a three-dimensional space that conforms to the actual environment. The path generation module 803, combined with preset obstacle avoidance constraints, generates a preliminary flight path that is more reasonable and safer. Thirdly, the segmentation module 804 divides the preliminary flight path into segments and extracts segment feature parameters, achieving a refined decomposition of the path, allowing the generation of control parameters to be specifically adapted to the path characteristics of different segments. Finally, the parameter generation module 805 generates preliminary control parameters based on the segment feature parameters, and the command generation module 806 integrates the preliminary flight path and preliminary control parameters into flight control commands, achieving deep binding between the flight path and control parameters. This invention generates flight control commands adapted to actual flight needs through a complete process of scene perception, three-dimensional modeling, path generation, segment feature extraction, parameter matching, and command integration, solving the problem of poor adaptability between flight paths and control parameters in the prior art and improving the flight stability and accuracy of UAVs.

[0044] Figure 9 This is a schematic diagram of the structure of a drone control device 900 provided in an embodiment of the present invention. The drone control device 900 can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 910 (e.g., one or more processors) and a memory 920, and one or more storage media 930 (e.g., one or more mass storage devices) for storing application programs 933 or data 932. The memory 920 and storage media 930 can be temporary or persistent storage. The program stored in the storage media 930 may include one or more modules (not shown in the diagram), each module may include a series of instruction operations on the drone control device 900. Furthermore, the processor 910 may be configured to communicate with the storage media 930 and execute the series of instruction operations in the storage media 930 on the drone control device 900 to implement the steps of the drone control methods provided in the above-described method embodiments.

[0045] The drone control device 900 may also include one or more power supplies 940, one or more wired or wireless network interfaces 950, one or more input / output interfaces 960, and / or one or more operating systems 931, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 9The structure of the drone control device shown does not constitute a limitation on the drone control device, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0046] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of the unmanned aerial vehicle control method.

[0047] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0048] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0049] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for controlling an unmanned aerial vehicle (UAV), characterized in that, The drone control method includes the following steps: Acquire the target flight scene and collect scene data from the target flight scene to obtain a scene dataset; 3D modeling is performed based on the scene dataset to obtain a 3D scene model; A preliminary flight path is generated based on a 3D scene model and preset obstacle avoidance constraints. The preliminary flight path is divided into segments to obtain segment characteristic parameters; Generate preliminary control parameters based on paragraph feature parameters; The initial flight path and initial control parameters are integrated into flight control commands.

2. The UAV control method according to claim 1, characterized in that, The process of acquiring the target flight scene and collecting scene data from the target flight scene to obtain a scene dataset includes: Acquire the target flight scene and determine the start and end points of the flight mission based on the target flight scene to define the data collection range; Static environmental data is collected based on the data collection range to obtain terrain data and obstacle data; Dynamic constraint data acquisition is performed based on the data acquisition range to obtain airspace constraint data and wind speed baseline data; The terrain data, obstacle data, airspace constraint data, and wind speed baseline data are integrated into a scene dataset.

3. The UAV control method according to claim 2, characterized in that, The process of performing 3D modeling based on the scene dataset to obtain a 3D scene model includes: Extract terrain data, obstacle data, airspace constraint data, and wind speed baseline data from the scene dataset; The terrain layer, obstacle layer, and airspace layer are constructed based on terrain data, obstacle data, and airspace constraint data, respectively. The terrain layer, obstacle layer, and airspace layer are aligned and superimposed according to spatial coordinates to generate a three-dimensional base model; Wind speed parameters are labeled for each spatial region in the 3D base model using wind speed baseline data to obtain the 3D scene model.

4. The UAV control method according to claim 1, characterized in that, The generation of the preliminary flight path based on the 3D scene model and preset obstacle avoidance constraints includes: Determine the spatial boundaries for path search based on a 3D scene model; The feasible path region is generated based on the preset obstacle avoidance constraints and spatial boundaries. Perform a path search within the feasible path region to generate a preliminary flight path.

5. The UAV control method according to claim 1, characterized in that, The process of dividing the initial flight path into segments to obtain segment feature parameters includes: The initial flight path is divided into segments based on preset segmentation rules to obtain multiple flight segments; Identify the average slope, path curvature, and obstacle spacing of each flight segment to obtain the characteristic parameters of each flight segment; The characteristic parameters of each flight segment are integrated into segment characteristic parameters according to the order of the flight segments.

6. The UAV control method according to claim 1, characterized in that, The generation of preliminary control parameters based on paragraph feature parameters includes: Based on preset mapping rules, paragraph feature parameters are mapped to initial control parameters; The performance threshold parameters of the UAV are obtained, and the out-of-limit parts of the initial control parameters are numerically corrected based on the performance threshold parameters to obtain the preliminary control parameters.

7. The UAV control method according to claim 1, characterized in that, The process of integrating the preliminary flight path and preliminary control parameters into flight control commands includes: Taking flight segments as units, the initial control parameters of each segment are bound to the initial flight path, and the parameter differences between adjacent segments are calculated. For segments with differences exceeding a preset threshold, gradient compensation for speed and steering angle is performed to obtain the path parameter set. Command conversion is performed based on the path parameter set to obtain flight control commands.

8. A drone control device, characterized in that, include: The data acquisition module is used to acquire the target flight scene and collect scene data from the target flight scene to obtain a scene dataset; The model building module is used to perform 3D modeling based on the scene dataset to obtain a 3D scene model; The path generation module is used to generate a preliminary flight path based on the 3D scene model and preset obstacle avoidance constraints. The segmentation module is used to segment the preliminary flight path to obtain segment feature parameters; The parameter generation module is used to generate preliminary control parameters based on paragraph feature parameters. The command generation module is used to integrate the preliminary flight path and preliminary control parameters into flight control commands.

9. A drone control device, characterized in that, The unmanned aerial vehicle control device includes: a memory and at least one processor, wherein the memory stores instructions; At least one of the processors invokes the instructions in the memory to cause the drone control device to perform the steps of the drone control method as described in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the steps of the UAV control method as described in any one of claims 1-7.

Images