Low-altitude flight path planning methods and platforms

Abstract

This invention belongs to the field of path planning technology and discloses a method and platform for low-altitude flight path planning. It involves acquiring three-dimensional environmental data and aircraft performance parameters of the target flight area, constructing an environmental model, and determining turning radius constraints, climb capability constraints, and safety separation constraints. Based on the environmental model, candidate flight paths are generated, and the local flyable space range at each path node is determined. Recovery trajectories are constructed for each path node. A set of feasible recovery trajectories is determined based on the spatial connectivity of the recovery trajectories in the environmental model. Recovery difficulty parameters are determined based on the changing trend of the spatial safety margin of the recovery trajectories and the margin changes that satisfy the turning radius and climb capability constraints of the aircraft. The path recoverability index is determined by combining the continuous reachability of the recovery trajectory to the safe area. Finally, the optimized target flight path is obtained. This invention can improve the safety and reliability of flight paths in complex low-altitude environments.

Description

Technical Field

[0001] This invention relates to the field of path planning technology, specifically to a low-altitude flight path planning method and platform. Background Technology

[0002] With the increasing application of low-altitude aircraft in urban logistics, power line inspection, emergency rescue, and environmental monitoring, flight path planning technology for complex low-altitude environments has gradually become a research focus. Existing low-altitude flight path planning methods typically rely on environmental map information, using obstacle avoidance algorithms or graph search algorithms to generate feasible paths that satisfy spatial and dynamic constraints, and further optimize indicators such as path length, flight energy consumption, or flight time. However, these methods primarily focus on the path passability of the aircraft under normal operating conditions, lacking effective consideration for unexpected anomalies that may occur during path execution.

[0003] In actual low-altitude flight, aircraft may enter narrow, localized areas such as gaps between buildings, passageways under bridges, forest corridors, or valley spaces. While the path itself may meet obstacle avoidance requirements in these areas, abnormal situations such as loss of positioning signal, communication link interruption, sudden wind disturbances, or decreased flight performance can prevent the aircraft from safely escaping the current area due to limited turning radius, insufficient ascent space, or insufficient lateral escape space. This results in a reachable but unrecoverable risky path. Existing path planning methods typically do not assess the recoverability at path nodes, making it difficult to guarantee the aircraft's safe evacuation capability under abnormal conditions. This leads to potential safety hazards in complex low-altitude scenarios due to the path planning results.

[0004] Therefore, it is necessary to propose a low-altitude flight path planning method that can evaluate the recovery capability of each node of the flight path during the path planning process and constrain the path generation based on the recovery capability, so as to improve the safety and robustness of low-altitude flight paths. Summary of the Invention

[0005] In view of the above-mentioned shortcomings mentioned in the background art, the purpose of this invention is to provide a low-altitude flight path planning method and platform.

[0006] The first aspect of the present invention provides a low-altitude flight path planning method, the method comprising the following steps: S1, acquiring three-dimensional environmental data and aircraft performance parameters of a target flight area, constructing an environmental model containing spatial distribution information of obstacles, and determining turning radius constraints, climb capability constraints, and safety separation constraints based on the aircraft performance parameters; S2, generating candidate flight paths composed of multiple path nodes based on the environmental model, and determining the local flyable space range at each path node; S3, constructing recovery trajectories for each path node, including at least one of a retreat trajectory, an ascent trajectory, and a lateral departure trajectory; S4, for each path node, determining a set of feasible recovery trajectories based on the spatial connectivity of the recovery trajectories in the environmental model, determining recovery difficulty parameters in the set of feasible recovery trajectories based on the changing trend of the spatial safety margin of the recovery trajectories and the margin changing characteristics that satisfy the turning radius constraints and climb capability constraints of the aircraft, and determining the path recoverability index of each path node in conjunction with the continuous reachability of the recovery trajectory to a preset safe area; S5, filtering or optimizing the candidate flight paths based on the path recoverability index of each path node to determine the target flight path.

[0007] A second aspect of the present invention provides a low-altitude flight path planning platform, the platform comprising: an environment modeling module, used to acquire three-dimensional environmental data and aircraft performance parameters of a target flight area, construct an environment model containing spatial distribution information of obstacles, and determine turning radius constraints, climb capability constraints and safety interval constraints based on the aircraft performance parameters;

[0008] The path generation module is used to generate candidate flight paths consisting of multiple path nodes based on the environment model, and to determine the local flyable space range at each path node; the recovery trajectory construction module is used to construct recovery trajectories for each path node, including at least one of a retreat trajectory, an ascent trajectory, and a lateral departure trajectory; the recovery capability assessment module is used to determine a set of feasible recovery trajectories for each path node based on the spatial connectivity of the recovery trajectory in the environment model, and to determine recovery difficulty parameters based on the spatial safety margin change trend of the recovery trajectory and the margin change characteristics that satisfy the aircraft's turning radius constraint and climb capability constraint in the set of feasible recovery trajectories, and to determine the path recoverability index of each path node in combination with the continuous reachability of the recovery trajectory to the preset safe area; the path determination module is used to screen or optimize the candidate flight paths based on the path recoverability index of each path node to determine the target flight path.

[0009] A third aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in any of the preceding claims.

[0010] Compared with existing technologies, this invention introduces a recovery trajectory construction and path recoverability assessment mechanism in the low-altitude flight path planning process. Based on the generation of candidate flight paths, it determines a set of feasible recovery trajectories according to the spatial connectivity of the recovery trajectories in the environmental model. It also determines the path recoverability index by combining the spatial safety margin change trend of the recovery trajectory, the margin change characteristics of the aircraft's turning radius constraint and climb capability constraint, and the continuous reachability of the recovery trajectory to the safe area. As a result, the planned target flight path not only meets obstacle avoidance and dynamic constraints, but also has the ability to safely evacuate under abnormal conditions. It can effectively prevent the aircraft from entering accessible but unrecoverable restricted space areas, improve the safety, robustness and controllability of flight paths in complex low-altitude environments, and enhance the adaptability of the path planning results to situations such as positioning anomalies, communication interruptions and sudden disturbances. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of the overall process of a low-altitude flight path planning method disclosed in an embodiment of the present invention;

[0012] Figure 2 This is a schematic diagram of the system architecture of the technical solution of the present invention;

[0013] Figure 3 This is a schematic diagram of the structure of a low-altitude flight path planning platform disclosed in an embodiment of the present invention. Detailed Implementation

[0014] The following specific embodiments illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0015] Furthermore, the technical features involved in the different embodiments of this application described below can be combined with each other as long as they do not conflict with each other.

[0016] Please see Figure 2 The solution in this embodiment can be deployed on the server, edge computing node, or flight control system corresponding to the low-altitude flight mission planning platform to generate safe and recoverable flight paths for UAVs or other low-altitude aircraft. Optionally, the low-altitude flight mission planning platform can also provide functional interfaces such as 3D environment display, flight path planning, path risk assessment, and path optimization, providing a supporting environment for data input and result display of the present invention.

[0017] It should be understood that the present invention is not limited to urban low-altitude environments, but is also applicable to scenarios such as mountain patrol, logistics distribution or emergency flight missions.

[0018] Please see Figure 1 This invention provides a low-altitude flight path planning method, comprising the following steps: S1, acquiring three-dimensional environmental data and aircraft performance parameters of the target flight area, constructing an environmental model containing spatial distribution information of obstacles, and determining turning radius constraints, climb capability constraints, and safety separation constraints based on the aircraft performance parameters; In this embodiment, the low-altitude flight area typically contains various obstacles such as buildings, trees, power lines, and terrain undulations, therefore, it is necessary to model the spatial structure of the target flight area. The three-dimensional environmental data can be derived from at least one or more of the following: lidar scanning data, three-dimensional model data generated by photogrammetry, existing GIS model data, or historical flight mission record data.

[0019] Based on the aforementioned three-dimensional environmental data, a flight area environmental model can be constructed. This environmental model describes the spatial distribution of obstacles and the basic structure of the flyable space within the flight area. Specifically, the environmental model can be established by spatially discretizing the three-dimensional environmental data, for example, by dividing the flight area into a regular voxel grid or a three-dimensional raster structure, where each grid cell represents whether an obstacle occupies the corresponding spatial area.

[0020] Through the above processing, an environmental model containing information on the spatial distribution of obstacles can be obtained.

[0021] After establishing the environmental model, it is also necessary to obtain the aircraft's performance parameters. These performance parameters reflect the range of maneuverability the aircraft can achieve during flight. The aircraft performance parameters include at least: minimum turning radius, maximum climb capability, maximum descent capability, maximum deceleration, and cruise speed range.

[0022] It is understandable that different aircraft have different maneuverability, so it is necessary to determine flight constraints based on the aircraft's performance parameters. Among them: turning radius constraints are used to ensure that the curvature of the flight path is within the allowable range of the aircraft; climb capability constraints are used to limit the rate of altitude change between adjacent segments in the path; safety separation constraints are used to ensure that the aircraft maintains a preset safe distance from obstacles. For example, the safety separation can be determined according to the size of the aircraft, such as 1-3 times the maximum size of the aircraft.

[0023] Through the above processing, the turning radius constraint, climbing ability constraint, and safety interval constraint used for path planning can be determined.

[0024] S2, Based on the environment model, a candidate flight path consisting of multiple path nodes is generated, and the local flyable space range at each path node is determined. In this embodiment, after establishing the flight area environment model and determining the aircraft performance constraints, candidate flight paths can be generated based on the environment model. The candidate flight path is used to describe one or more feasible flight paths for the aircraft to fly from the starting position to the target position.

[0025] Understandably, candidate flight paths are used to reflect the paths that an aircraft can take under the conditions of avoiding obstacles and meeting basic flight constraints. However, these paths do not take into account the aircraft's evacuation capability in abnormal situations. Therefore, further recovery capability assessments of candidate flight paths are needed in subsequent steps.

[0026] A candidate flight path can be represented as a sequence of path nodes, where each node is connected sequentially according to the flight order to form a complete path. Each path node corresponds to the spatial position of the aircraft at a given moment. For example, a candidate flight path can be represented as: Path node sequence: P = {p1, p2, ..., p...} }. Where: p This represents the spatial coordinates of the i-th path node. Path nodes can be connected by straight line segments or curved line segments to form a continuous flight path.

[0027] As an example, generating candidate flight paths composed of multiple path nodes based on the environmental model and determining the local wingable space range at each path node includes: S21, performing voxelization processing on the three-dimensional environmental data to obtain an obstacle occupancy grid, and expanding the obstacle occupancy grid based on the safety interval constraint to obtain a wingable space grid; in this step, the flight area environmental model established in step S1 is further transformed into a spatial representation suitable for path planning calculation, so as to effectively describe the wingable space in the flight area. It is understood that three-dimensional environmental data usually exists in the form of point cloud data or three-dimensional models. Directly using the raw data for path planning calculation is inefficient; therefore, spatial discretization processing of the three-dimensional environmental data is necessary.

[0028] Specifically, the flight area is divided into a regular voxel grid structure, with each voxel unit corresponding to a three-dimensional spatial region of a fixed size. For example, in an urban low-altitude flight scenario, the voxel unit size can be set to 0.5-1 meter horizontally and 0.5-1 meter vertically. It is understandable that smaller voxel sizes result in higher spatial description accuracy, but also increase computational complexity. Therefore, an appropriate voxel size can be selected based on the required planning accuracy.

[0029] After completing the voxel division, the occupancy status of each voxel unit is determined based on the 3D environment data: when there is an obstacle point in a voxel unit, the voxel is marked as an obstacle-occupied voxel; otherwise, it is marked as a flyable voxel.

[0030] It is understandable that the voxels occupied by the obstacles mentioned above reflect the actual geometric boundaries of the obstacles. However, during actual flight, the aircraft needs to maintain a certain safe distance from the obstacles, thus requiring the introduction of safety interval constraints. The voxels occupied by the obstacles can be expanded based on these safety interval constraints, causing the obstacle boundaries to extend outwards by a preset distance. For example, when the safety interval is set to 2 meters and the voxel size is 0.5 meters, voxels within 4 voxels of the obstacle-occupying voxel can be marked as non-flying voxels.

[0031] After the above expansion process, the path generated by the aircraft in the flyable space will always maintain a safe distance from obstacles, thereby improving flight safety. The expansion process can be implemented through a neighborhood search method, that is, for each obstacle occupying a voxel, voxels within a certain range around it are marked as non-flyable voxels.

[0032] After the above processing, a flyable space grid can be obtained. The flyable space grid is used to describe the distribution of flyable and non-flyable space in the flight area and serves as the basis for subsequent path planning.

[0033] S22, in the flyable space grid, an initial path is generated based on the starting and ending positions using a graph search algorithm or a sampling planning algorithm, and the initial path is subjected to waypoint thinning and / or smoothing to obtain a candidate flight path composed of multiple path nodes; in this step, a candidate path for the aircraft to fly from the starting position to the target position is generated in the aforementioned flyable space grid.

[0034] Specifically, the position coordinates of the flight start and end points are determined based on the flight mission. The start point is the aircraft's takeoff position or current location, and the end point is the mission target location. An initial flight path can then be generated using a path search algorithm within the flyable space grid.

[0035] Specifically, graph search algorithms, such as the A* algorithm or Dijkstra's algorithm, can be used for path planning. When using a graph search algorithm, each flyable voxel can be considered a graph node, and connections are established between adjacent voxels to form a spatial graph structure. A connected path from the starting voxel to the ending voxel can then be obtained through the graph search algorithm. Alternatively, sampling planning algorithms, such as the RRT algorithm or PRM algorithm, can also be used to generate paths.

[0036] Understandably, initial paths typically consist of a large number of voxel nodes, and the path direction changes frequently, which is detrimental to the actual flight control of the aircraft. Therefore, the initial path needs further processing as follows: waypoint thinning is performed on the initial path to maintain a reasonable distance between path nodes. For example, when the distance between adjacent path nodes is less than a preset distance threshold, one of the path nodes can be deleted. Understandably, waypoint thinning can reduce the number of path nodes, thereby reducing the computational burden on the flight control system.

[0037] Simultaneously, the path can be smoothed to make changes in path direction more continuous. Curve fitting methods can be used for path smoothing, such as fitting a spline curve to the path node sequence. Path smoothing avoids excessive directional changes, thus meeting the aircraft's turning capability constraints.

[0038] After the above processing, a candidate flight path consisting of multiple path nodes can be obtained.

[0039] S23. For any path node in the candidate flight path, extract the connected flyable region in a preset neighborhood with the path node as the center, and trim the connected flyable region in combination with the turning radius constraint and the climb capability constraint to obtain the local flyable space range at the path node.

[0040] In this step, it is necessary to determine the local flyable space range near each path node in the candidate flight path. Understandably, the spatial structure of different path nodes varies greatly. For example, some path nodes may be located in open areas, while others may be located in narrow passages between buildings. Therefore, it is necessary to analyze the flyable space structure near each path node separately.

[0041] Specifically, a pre-defined neighborhood space is established centered on the path node. For example, a spherical or cubic neighborhood can be established centered on the path node, and its scale can be determined according to the aircraft's performance parameters. For example, the neighborhood radius can be set to 1-3 times the aircraft's minimum turning radius.

[0042] After establishing the neighborhood space, a set of flyable voxels connected to the voxel containing the path node is extracted within this neighborhood. This set represents the spatial region that the aircraft can enter near the path node. It is understood that if some flyable voxels are located within the neighborhood but are not connected to the voxel containing the path node, the aircraft cannot actually enter that region and therefore should not be included in the local flyable space.

[0043] After extracting the connected flyable region, it is necessary to trim this region in conjunction with aircraft performance constraints. Specifically, the turning radius constraint ensures sufficient space near path nodes for the aircraft to complete turning maneuvers. For example, if the channel width in a certain direction is less than the spatial scale corresponding to the minimum turning radius, it can be assumed that the aircraft cannot complete a turning maneuver in that direction, and therefore the corresponding region can be removed from the local flyable space.

[0044] At the same time, climb capability constraints also need to be considered. For example, if the rate of altitude change in a certain direction exceeds the aircraft's allowable climb capability, it can be assumed that the aircraft cannot complete flight maneuvers in that direction, and therefore the corresponding area can be excluded from the local flyable space.

[0045] Through the above trimming process, the local flyable space range at each path node can be obtained.

[0046] S3, construct recovery trajectories for each path node, including at least one of a retreat trajectory, an ascent trajectory, and a lateral departure trajectory. In this embodiment, after obtaining the candidate flight paths and the local flyable space range at each path node, recovery trajectories are further constructed for each path node. It is understood that during low-altitude flight, the aircraft may deviate from its original planned path due to positioning errors, communication anomalies, sudden disturbances, or control anomalies. In this case, if there is no feasible evacuation path near the path node, the aircraft may enter a space area that is reachable but difficult to evacuate safely, thereby increasing flight risk. Therefore, in this step, recovery trajectories are constructed for each path node in the candidate flight paths to describe the possible flight trajectories that allow the aircraft to leave the current path area when an anomaly occurs at a path node.

[0047] Understandably, different path nodes are situated in different environmental structures. For example, some path nodes may be located in open areas, while others may be located in narrow passages between buildings. Therefore, it is necessary to construct recovery trajectories for each path node separately. Recovery trajectories can include at least one of the following: a retreat trajectory, an ascent trajectory, or a lateral disengagement trajectory. These recovery trajectories correspond to different evacuation methods and can be used to describe various evacuation possibilities for the aircraft at path nodes.

[0048] As an example, a recovery trajectory is constructed for each path node, including: for any path node, extracting a path segment corresponding to a preset backtracking distance along the reverse direction of the candidate flight path, and superimposing a braking distance determined by the flight speed and maximum deceleration at the path node into the backtracking distance to generate a backtracking trajectory; specifically, when an abnormal situation occurs at a path node, the most direct evacuation method is to fly in the reverse direction along the original flight path to a safe area, so a backtracking trajectory is usually highly feasible. Path segments within a preset backtracking distance range can be extracted along the reverse direction of the candidate flight path as the base path segments of the backtracking trajectory. The preset backtracking distance can be determined according to the aircraft's flight speed. For example, at higher flight speeds, the backtracking distance can be appropriately increased to ensure that the aircraft has sufficient space to complete deceleration and turning maneuvers.

[0049] Understandably, an aircraft typically undergoes a deceleration process before flying along its retreat trajectory. Therefore, considering only the geometric retreat distance may be insufficient to ensure flight safety, necessitating further consideration of the braking distance. The braking distance can be determined based on the aircraft's speed at path nodes and its maximum deceleration. For example, when the aircraft's speed is 10 m / s and its maximum deceleration is 2 m / s², the braking distance can be determined accordingly. 2 At that time, the braking distance is approximately 25 meters. Therefore, when extracting the reversal path segment, the braking distance can be superimposed on the preset reversal distance.

[0050] Understandably, by introducing braking distance, the aircraft can have enough space to complete the deceleration process when executing the reversal trajectory, thereby improving the feasibility of the reversal trajectory.

[0051] For any path node, an ascent segment is constructed along the altitude direction from the path node to a preset safe altitude, and an ascent safety body considering the safety interval constraint is constructed around the ascent segment. An ascent trajectory is generated based on the collision detection results between the ascent safety body and the environment model. Specifically, in a low-altitude flight environment, the ascent trajectory can usually allow the aircraft to get away from areas with dense ground obstacles, thus it is an effective evacuation method.

[0052] An ascent segment can be established vertically, starting from a path node. The endpoint height of the ascent segment can be set to a preset safe height. The preset safe height can be determined based on the environmental structure of the flight area. For example, in densely built-up areas, the safe height can be set to a preset distance above the height of the highest obstacle in the area.

[0053] After constructing the ascent section, the safe distance between the aircraft and surrounding obstacles needs to be considered. Therefore, an ascent safety body can be established around the ascent section, which is represented as a cylindrical spatial region with the ascent section as its central axis, and its radius can be equal to the safe separation constraint distance.

[0054] The next step is to determine whether the ascending safety body overlaps with obstacles in the environment model. Understandably, if the ascending safety body overlaps with an obstacle, it indicates a potential collision risk during the ascent trajectory, making that trajectory infeasible. An ascent trajectory can be generated when the ascending safety body does not overlap with an obstacle.

[0055] For any path node, a lateral departure direction is determined based on the heading direction of the path node. A target departure area that meets a preset openness threshold is searched in the lateral departure direction. A lateral departure trajectory connecting the path node and the target departure area is generated under the condition of satisfying the turning radius constraint.

[0056] Specifically, when an aircraft is located in a narrow passage or between buildings, retreating or ascending along the path may not be a safe evacuation. In this case, it can enter a nearby open area by laterally deviating from the trajectory.

[0057] The lateral departure direction can be determined first based on the flight direction at the path node. For example, the left or right direction can be determined as the lateral departure direction based on the path direction.

[0058] The target departure area can then be searched in the lateral departure direction. The target departure area can be defined as a flyable space region that meets a preset openness threshold. For example, when the flyable space volume or minimum passage width of a certain area exceeds the preset threshold, that area can be considered the target departure area.

[0059] After determining the target departure area, a flight trajectory connecting the path nodes to the target departure area can be generated under the condition of satisfying the turning radius constraint. It can be understood that by satisfying the turning radius constraint, it can be ensured that the aircraft can complete the trajectory turning maneuver during actual flight.

[0060] S4. For each path node, a set of feasible recovery trajectories is determined based on the spatial connectivity of the recovery trajectory in the environmental model. Within this set, recovery difficulty parameters are determined based on the spatial safety margin variation trend of the recovery trajectory and the margin variation characteristics that satisfy the aircraft's turning radius and climb capability constraints. Furthermore, the path recoverability index for each path node is determined by combining the continuous reachability of the recovery trajectory to the preset safe area. In this embodiment, although multiple recovery trajectories have been constructed, different recovery trajectories may face spatial limitations or obstacles in the actual environment. Therefore, it is necessary to further determine the spatial connectivity of the recovery trajectories to select truly feasible recovery trajectories.

[0061] Furthermore, even if a recovery trajectory is spatially connected, its safety margin and flight difficulty may vary significantly. For example, some recovery trajectories may be feasible, but their safety margins are small or the aircraft's motion constraint margins are low, resulting in higher risks in actual flight. Therefore, further assessment of the recovery trajectory's difficulty is necessary.

[0062] In this embodiment, by comprehensively analyzing the spatial connectivity of the recovery trajectory, the changing trend of the spatial safety margin, and the changing characteristics of the aircraft constraint margin, and combining this with the continuous accessibility of the recovery trajectory to the safe area, the path recoverability index of each path node can be determined.

[0063] As an example, for each path node, a set of feasible recovery trajectories is determined based on the spatial connectivity of the recovery trajectory in the environment model, including: S41, discretizing each recovery trajectory into multiple sampling points along the trajectory direction, and constructing a local safety domain that satisfies the safety interval constraint at each sampling point; in this step, the recovery trajectory can be represented as a continuous spatial curve (e.g., a retreat trajectory, an ascending trajectory, or a lateral departure trajectory). In order to facilitate the determination of the spatial relationship between the recovery trajectory and the environment model, it is necessary to discretize the recovery trajectory into multiple sampling points.

[0064] Let a certain recovered trajectory be denoted as Its arc length is Along the trajectory arc length direction at sampling intervals Discretize the samples to obtain the set of sampling points: ;in: . A value of 0.5m-2m is chosen to balance the accuracy of the judgment with the efficiency of the calculation.

[0065] At each sampling point Constructing a local security domain Used to reflect safety interval constraints The safe zone occupied by the aircraft. The local safety zone can be constructed as a spherical region centered on the sampling point: .

[0066] Of course, it can also be constructed as a cylindrical or ellipsoidal domain to better reflect the shape and attitude changes of the aircraft. However, regardless of the form used, the minimum circumscribed radius of the local safety domain should at least be no less than [a certain value]. This is to ensure the consistency of safety intervals.

[0067] S42, based on the intersection relationship between the local safety domain and the obstacle-occupied area in the environment model, determine the spatial connectivity parameters of each recovery trajectory, including collision markers and continuous safety connectivity length; in this step, based on obtaining the set of sampling points corresponding to each recovery trajectory and the local safety domain at each sampling point, it is necessary to further determine the spatial connectivity of the recovery trajectory according to the spatial relationship between the local safety domain and the obstacle-occupied area in the environment model.

[0068] It is understandable that if the local safety domain corresponding to any sampling point on the recovery trajectory overlaps with the area occupied by the obstacle, it means that the aircraft may be at risk of collision when executing the recovery trajectory, and therefore the recovery trajectory cannot be safely passed through that location.

[0069] Let Ω be the local safety region corresponding to the j-th sampling point on the k-th recovered trajectory. In the environment model, the area occupied by obstacles is The collision state at the sampling point can be determined by judging whether there is an overlap between the two: Where: c =1 indicates that there is a collision risk at this sampling point; c =0 indicates that the safety interval constraint is met at this sampling point.

[0070] After obtaining the collision status of each sampling point, the collision markers of the recovered trajectory can be further determined: .

[0071] Understandably, when the collision marker C... A value of 1 indicates that there is at least one sampling point on the recovered trajectory at risk of collision; when the collision flag C... When the value is 0, it indicates that the recovered trajectory has no collisions in the sampling sense.

[0072] After identifying the collision markers, it is necessary to further determine the continuous safe continuity length of the recovery trajectory. Understandably, the continuous safe continuity length describes the maximum distance the recovery trajectory can safely traverse continuously from the path nodes; this length reflects the spatial range within which the aircraft can safely fly along the recovery direction.

[0073] The sampling point number where the first collision occurs on the recovered trajectory can be represented as: When there are no collision sampling points, the recovered trajectory can be considered to be safely continuous. Therefore, the continuous safe continuity length of the recovered trajectory can be expressed as: Where: Δs is the sampling interval; L To restore the total length of the trajectory.

[0074] Through the above processing, the spatial connectivity parameters of the recovered trajectory can be obtained, including the collision marker C. and continuous safe penetration length .

[0075] S43, Filter the recovery trajectory according to the spatial connectivity parameters: when the collision marker indicates no collision and the continuous safe connectivity length is not less than the preset connectivity threshold, retain the corresponding recovery trajectory; in this step, when there is a collision risk in the recovery trajectory, the aircraft may collide during the execution of the recovery trajectory, so the recovery trajectory should not be retained as a feasible recovery trajectory.

[0076] Furthermore, even if the recovery trajectory does not pose a collision risk in space, if the continuous safe length is too short, the aircraft may still be unable to complete a safe evacuation during the execution of the recovery trajectory. Therefore, such a recovery trajectory is also not suitable as a feasible recovery trajectory.

[0077] Therefore, a preset penetration threshold L can be set. th This characterizes the minimum continuous, safe penetration length that the recovered trajectory should have. For example, the penetration threshold can be determined based on the aircraft's braking distance: Where: v is the flight speed at the path node; a max This is the maximum deceleration of the aircraft.

[0078] Understandably, by ensuring that the continuous safe penetration length is not less than the braking distance, it can be guaranteed that the aircraft has sufficient space to complete deceleration and attitude adjustment actions when performing the recovery trajectory.

[0079] When the above conditions are met, the recovered trajectory can be considered to have sufficient spatial connectivity. Therefore, the criteria for selecting the recovered trajectory can be expressed as: Among them: Γ Indicates the recovered trajectory; This represents the set of feasible recoverable trajectories corresponding to path nodes.

[0080] S44, collect the retained recovery trajectories to obtain a set of feasible recovery trajectories corresponding to the path nodes.

[0081] In this step, the aforementioned has already been based on the collision marker. and continuous safe penetration length The recovered trajectories were screened, and the retained trajectories all met the spatial connectivity conditions, so they can be used as feasible evacuation paths at the path nodes.

[0082] For any path node The set of feasible recovery trajectories corresponding to it can be represented as: ;in: This represents the recovery trajectory corresponding to the path node; Collision markers indicating the recovery trajectory; Indicates the continuous safe penetration length of the restored trajectory; This indicates the preset penetration threshold.

[0083] Through the above processing, a set of feasible recovery trajectories corresponding to each path node can be obtained. The set of feasible recovery trajectories is used for subsequent calculation of recovery difficulty parameters and determination of path recoverability indicators.

[0084] As an example, the recovery difficulty parameter is determined based on the spatial safety margin change trend of the recovery trajectory and the margin change characteristics that satisfy the aircraft's turning radius constraint and climb capability constraint in the set of feasible recovery trajectories. The path recoverability index of each path node is determined in conjunction with the continuous reachability of the recovery trajectory to the preset safe area. This includes: S45, sampling the recovery trajectories in the set of feasible recovery trajectories along the trajectory direction, calculating the spatial safety margin at each sampling point, and determining the minimum spatial safety margin and the rate of decrease of the safety margin as spatial safety margin change trend parameters from the spatial safety margin sequence corresponding to each recovery trajectory. In this step, by sampling the selected feasible recovery trajectories, the spatial safety margin between each sampling point and the obstacle (i.e., the minimum distance between the aircraft and the obstacle) is calculated, and the change trend of this safety margin is quantitatively analyzed.

[0085] Suppose a certain recovery trajectory The j-th sampling point is Calculate the spatial safety margin at the sampling point. That is, the minimum distance between the point and the nearest obstacle minus the safety interval. : ;in: Sampling points Distance to the nearest obstacle; This refers to the safe distance between the aircraft and the obstacle.

[0086] The spatial safety margin sequence is calculated from all sampling points along the restored trajectory: ,in This represents the spatial safety margin at the j-th sampling point.

[0087] Minimum safety margin for space This represents the minimum spatial safety margin at a path node, used to measure the minimum distance between the aircraft and obstacles. Its calculation formula is: .

[0088] Safety margin decline rate This represents the rate at which the safety margin between the aircraft and obstacles on the recovery trajectory changes as the path advances. A higher descent rate indicates a faster entry into the danger zone, making recovery more difficult. The calculation formula is: ;in To recover the total length of the trajectory, and These are the spatial safety margins for the starting and ending points of the trajectory, respectively.

[0089] Understandably, by calculating the minimum safety margin and the rate of decrease of the safety margin, the safety of the recovery trajectory in space can be quantitatively assessed. The larger the minimum safety margin, the farther the aircraft is from the obstacle, and the higher the safety; while the larger the rate of decrease, the faster the safety margin decreases during the execution of the recovery trajectory, and the higher the flight difficulty and risk.

[0090] S46, calculate the margin variation parameters for the recovery trajectory to satisfy the turning radius constraint and climb capability constraint. Based on the spatial safety margin variation trend parameter and the margin variation parameters corresponding to the recovery trajectory, determine the recovery difficulty parameters corresponding to each recovery trajectory. In this step, it is first necessary to calculate the margin variation parameters of the recovery trajectory, including the turning margin variation and the climb margin variation. These two margin variation parameters reflect the operational difficulty of the aircraft in the recovery trajectory.

[0091] Changes in turning margin For each sampling point on the recovered trajectory Calculate the turning radius at that point. The change in turn margin indicates whether an aircraft has enough room to complete a turn, determined by the aircraft's minimum turning radius. Limitations. If Less than If the error is negative, it means the aircraft cannot complete the turn smoothly on that path segment, resulting in a negative turn margin and increased difficulty in recovery.

[0092] The calculation formula is: .

[0093] when When the time is right, it indicates that the path segment presents a significant turning challenge.

[0094] Climb margin changes For each sampling point on the recovered trajectory Calculate the rate of ascent at that point. Then, with the aircraft's maximum climb rate Comparisons are made. Changes in climb margin reflect the operational difficulty of the aircraft during its climb along the path. If the climb rate is close to the aircraft's maximum climb capability, the climb margin is small, and the recovery becomes more difficult.

[0095] The calculation formula is: .

[0096] when When the value is small or negative, it indicates that the aircraft's climb ability is limited, and recovery becomes more difficult.

[0097] Next, the safety of the path is further evaluated based on the trend parameter of the spatial safety margin change of the recovered trajectory.

[0098] Minimum safety margin for space This is the minimum safety margin between the aircraft and obstacles in the recovery trajectory. This value reflects the safety of the aircraft on the recovery path; a smaller safety margin means that the aircraft faces a higher risk.

[0099] Safety margin decline rate This indicates the rate at which the safety margin between the aircraft and obstacles on the recovery trajectory changes as the aircraft progresses along the trajectory. A larger descent rate indicates that the aircraft is approaching the danger zone at a faster speed, and the recovery trajectory is less safe.

[0100] Finally, based on the margin variation parameters and the space safety margin variation trend parameters, the recovery difficulty parameters of the recovery trajectory are determined. This parameter quantifies the difficulty of recovering path nodes; larger ones... This indicates that the path node recovery is difficult, and the aircraft faces significant challenges and risks in executing this trajectory.

[0101] Recovery Difficulty Parameters Based on the spatial safety margin variation trend parameter (reflecting the safety of path nodes; the smaller the spatial safety margin, the higher the recovery difficulty) and the margin variation parameter (changes in turning margin and climb margin reflect the operational difficulty of the aircraft in the recovery trajectory; the greater the margin change, the higher the recovery difficulty), the recovery difficulty parameter can be expressed as: ;in: and It reflects the spatial safety margin of path nodes and its changes; and This reflects the changes in the turning margin and climb margin of the recovered trajectory.

[0102] It is understood that the above function f() can be a weighted summation function (that is, the space safety margin, descent rate, turning margin change and climb margin change are normalized and then weighted and summed) or a multiplication function (that is, the space safety margin, descent rate, turning margin change and climb margin change are normalized and then multiplied together), and its specific calculation expression is not specifically limited.

[0103] S47. Based on the continuous reachable path length from the path node to the preset safe area and the corresponding recovery difficulty parameter, determine the trajectory recoverability value of each recovery trajectory, and select the maximum value of the trajectory recoverability value as the path recoverability index of the path node.

[0104] In this step, the path recoverability metric is used to quantify the recoverability of path nodes. By comprehensively considering the length of the continuously reachable path of the recovered trajectory and the recovery difficulty, the recoverability of each path node is determined.

[0105] For each recovery trajectory Calculate from path node Start by restoring the trajectory to the length of the continuously reachable path to the preset safe area. The path length is determined by checking whether the sampling points of the recovered trajectory enter the safe zone. To determine: .

[0106] The sampling point number for the first entry into the safe area is The length of the continuously reachable path is: .

[0107] The recoverability value of the trajectory is calculated based on the continuous reachable path length and recovery difficulty parameters. : ;in To restore the difficulty parameters, if If it is larger, then A smaller value indicates that the trajectory is more difficult to recover and has a lower recoverability value.

[0108] For path nodes Select the trajectory recoverability value from all feasible recoverable trajectories. The maximum value is used as the path recoverability index for that path node: .

[0109] When the set of feasible recovery trajectories is empty It can be set to 0 to indicate that the path node is unrecoverable.

[0110] S5. Based on the path recoverability index of each path node, the candidate flight paths are screened or optimized to determine the target flight path.

[0111] In this step, candidate flight paths are screened or optimized based on the path recoverability index of each path node, thereby determining the optimal target flight path.

[0112] First, a weighted sum of the path recoverability indices for all path nodes is calculated. A higher path recoverability index means a stronger recovery capability for that path node, and a greater likelihood that the aircraft can safely evacuate if an anomaly occurs at that node. Therefore, the weighted sum reflects the safety and reliability of the entire flight path. Understandably, the importance of different path nodes can be adjusted through weights during the calculation. For example, the start and end points of the path typically have higher weights because they are crucial to the successful execution of the path.

[0113] Based on a path recoverability weighted sum, candidate flight paths are further optimized. The optimization objective is to maximize path recoverability while considering flight path length or other mission requirements.

[0114] During optimization, the path with the highest recoverability weighted sum is prioritized to ensure the aircraft has maximum recovery capability in abnormal situations. Additionally, the length of the flight path may be appropriately controlled to improve flight efficiency.

[0115] After optimization, a flight path that meets the aircraft's operational requirements is selected as the target flight path. This ensures that the path not only meets recovery requirements but also conforms to the aircraft's performance constraints and the actual needs of the flight mission. For example, the flight path needs to meet constraints such as the aircraft's turning radius, maximum climb capability, and safe separation, and the path length should be within the time limit required by the mission.

[0116] Finally, the selected target flight path is output and transmitted to the flight control system. The flight control system guides the aircraft to complete the mission according to this path, ensuring that the aircraft can successfully execute the flight mission along the predetermined route.

[0117] Please see Figure 3This invention also provides a low-altitude flight path planning platform 100, comprising: an environment modeling module 10, used to acquire three-dimensional environmental data and aircraft performance parameters of a target flight area, construct an environment model containing spatial distribution information of obstacles, and determine turning radius constraints, climb capability constraints, and safety interval constraints based on the aircraft performance parameters; a path generation module 20, used to generate candidate flight paths composed of multiple path nodes based on the environment model, and determine the local flyable space range at each path node; and a recovery trajectory construction module 30, used to construct recovery trajectories for each path node, including a retreat trajectory, an ascent trajectory, and a recovery trajectory. At least one of the lateral departure trajectories; recovery capability assessment module 40, used to determine a set of feasible recovery trajectories for each path node based on the spatial connectivity of the recovery trajectory in the environmental model, and to determine recovery difficulty parameters based on the spatial safety margin change trend of the recovery trajectory and the margin change characteristics that satisfy the turning radius constraint and climb capability constraint of the aircraft in the set of feasible recovery trajectories, and to determine the path recoverability index of each path node in combination with the continuous reachability of the recovery trajectory to the preset safe area; path determination module 50, used to screen or optimize the candidate flight paths according to the path recoverability index of each path node, and determine the target flight path.

[0118] As an example, the path generation module 20 is specifically used to: perform voxelization processing on the three-dimensional environment data to obtain an obstacle-occupying grid, and perform dilation processing on the obstacle-occupying grid based on the safety interval constraint to obtain a flyable space grid; in the flyable space grid, generate an initial path using a graph search algorithm or a sampling planning algorithm based on the start and end positions, and perform waypoint thinning and / or smoothing processing on the initial path to obtain a candidate flight path composed of multiple path nodes; for any path node in the candidate flight path, extract a connected flyable region in a preset neighborhood with the path node as the center, and trim the connected flyable region in combination with the turning radius constraint and the climb capability constraint to obtain the local flyable space range at the path node.

[0119] As an example, the recovery trajectory construction module 30 is specifically used to: for any path node, extract the path segment corresponding to the preset backtracking distance along the reverse of the candidate flight path, and superimpose the braking distance determined by the flight speed and maximum deceleration at the path node into the backtracking distance to generate a backtracking trajectory; for any path node, construct an ascent segment from the path node along the altitude direction to a preset safe altitude, and construct an ascent safety body considering the safety interval constraint around the ascent segment, and generate an ascent trajectory based on the collision detection results of the ascent safety body and the environment model; for any path node, determine the lateral departure direction based on the heading direction of the path node, search for a target departure area that meets a preset openness threshold in the lateral departure direction, and generate a lateral departure trajectory connecting the path node and the target departure area under the condition of meeting the turning radius constraint.

[0120] As an example, the recovery capability assessment module 40 is specifically used to: discretize each recovery trajectory into multiple sampling points along the trajectory direction, and construct a local safety domain that satisfies the safety interval constraint at each sampling point; determine the spatial connectivity parameters of each recovery trajectory based on the intersection relationship between the local safety domain and the obstacle-occupied area in the environment model, including collision markers and continuous safe connectivity lengths; filter recovery trajectories according to the spatial connectivity parameters: when the collision marker indicates no collision and the continuous safe connectivity length is not less than a preset connectivity threshold, retain the corresponding recovery trajectory; and aggregate the retained recovery trajectories to obtain a set of feasible recovery trajectories corresponding to the path nodes.

[0121] As an example, the recovery capability assessment module 40 is further configured to: sample the recovery trajectories in the set of feasible recovery trajectories along the trajectory direction, calculate the spatial safety margin at each sampling point, and determine the minimum spatial safety margin and the rate of decrease of the safety margin as spatial safety margin change trend parameters based on the spatial safety margin sequence corresponding to each recovery trajectory; calculate the margin change parameters for the recovery trajectory to satisfy the turning radius constraint and the climbing capability constraint, and determine the recovery difficulty parameter corresponding to each recovery trajectory based on the spatial safety margin change trend parameter and the margin change parameter corresponding to the recovery trajectory; determine the trajectory recoverability value of each recovery trajectory based on the continuous reachable path length from the path node to the preset safe area and the corresponding recovery difficulty parameter, and select the maximum value of the trajectory recoverability value as the path recoverability index of the path node.

[0122] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in any of the preceding embodiments.

[0123] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A low-altitude flight path planning method, characterized in that, The method includes the following steps: S1, acquiring three-dimensional environmental data and aircraft performance parameters of the target flight area, constructing an environmental model containing spatial distribution information of obstacles, and determining turning radius constraints, climb capability constraints, and safety interval constraints based on the aircraft performance parameters; S2, generating candidate flight paths composed of multiple path nodes based on the environmental model, and determining the local flyable space range at each path node; S3, constructing recovery trajectories for each path node, including at least one of a retreat trajectory, an ascent trajectory, and a lateral departure trajectory; S4, for each path node, determining a set of feasible recovery trajectories based on the spatial connectivity of the recovery trajectory in the environmental model, determining recovery difficulty parameters based on the spatial safety margin change trend of the recovery trajectory and the margin change characteristics that satisfy the turning radius constraints and climb capability constraints of the aircraft in the set of feasible recovery trajectories, and determining the path recoverability index of each path node in combination with the continuous reachability of the recovery trajectory to the preset safety area; S5, filtering or optimizing the candidate flight paths based on the path recoverability index of each path node to determine the target flight path. For each path node, a set of feasible recovery trajectories is determined based on the spatial connectivity of the recovery trajectory in the environment model, including: S41, discretizing each recovery trajectory into multiple sampling points along the trajectory direction, and constructing a local safety domain satisfying the safety interval constraint at each sampling point; S42, determining the spatial connectivity parameters of each recovery trajectory based on the intersection relationship between the local safety domain and the obstacle-occupied area in the environment model, including collision markers and continuous safe connectivity length; S43, filtering recovery trajectories according to the spatial connectivity parameters: when the collision marker indicates no collision and the continuous safe connectivity length is not less than a preset connectivity threshold, the corresponding recovery trajectory is retained; S44, aggregating the retained recovery trajectories to obtain a set of feasible recovery trajectories corresponding to the path node. The recovery difficulty parameter is determined based on the spatial safety margin change trend and the margin change characteristics that satisfy the aircraft's turning radius constraint and climb capability constraint in the set of feasible recovery trajectories. The path recoverability index of each path node is determined in conjunction with the continuous reachability of the recovery trajectory to the preset safe area. This includes: S45, sampling the recovery trajectories in the set of feasible recovery trajectories along the trajectory direction, calculating the spatial safety margin at each sampling point, and determining the minimum spatial safety margin and the safety margin descent rate as spatial safety margin change trend parameters from the spatial safety margin sequence corresponding to each recovery trajectory; S46, calculating the margin change parameters that satisfy the turning radius constraint and climb capability constraint for the recovery trajectory, and determining the recovery difficulty parameter corresponding to each recovery trajectory based on the spatial safety margin change trend parameter and the margin change parameter; S47, determining the trajectory recoverability value of each recovery trajectory based on the continuous reachability path length from the path node to the preset safe area and the corresponding recovery difficulty parameter, and selecting the maximum value of the trajectory recoverability value as the path recoverability index of the path node.

2. The low-altitude flight path planning method according to claim 1, characterized in that: The process involves generating candidate flight paths composed of multiple path nodes based on the environmental model, and determining the local flyable space range at each path node, including: S21, performing voxelization on the three-dimensional environmental data to obtain an obstacle-occupied grid, and expanding the obstacle-occupied grid based on the safety interval constraint to obtain a flyable space grid; S22, generating an initial path in the flyable space grid using a graph search algorithm or a sampling planning algorithm based on the start and end positions, and performing waypoint thinning and / or smoothing on the initial path to obtain candidate flight paths composed of multiple path nodes; S23, for any path node in the candidate flight paths, extracting a connected flyable region within a preset neighborhood centered on the path node, and cropping the connected flyable region in conjunction with the turning radius constraint and the climb capability constraint to obtain the local flyable space range at the path node.

3. The low-altitude flight path planning method according to claim 1, characterized in that: For each path node, a recovery trajectory is constructed, including: for any path node, extracting a path segment corresponding to a preset backtracking distance along the reverse of the candidate flight path, and superimposing a braking distance determined by the flight speed and maximum deceleration at the path node into the backtracking distance to generate a backtracking trajectory; for any path node, constructing an ascent segment from the path node along the altitude direction to a preset safe altitude, and constructing an ascent safety body considering the safety interval constraint around the ascent segment, generating an ascent trajectory based on the collision detection results between the ascent safety body and the environment model; for any path node, determining a lateral departure direction based on the heading direction of the path node, searching for a target departure area that meets a preset openness threshold in the lateral departure direction, and generating a lateral departure trajectory connecting the path node and the target departure area under the condition of meeting the turning radius constraint.

4. A low-altitude flight path planning platform, characterized in that: The platform includes: an environment modeling module, used to acquire three-dimensional environmental data and aircraft performance parameters of the target flight area, construct an environment model containing spatial distribution information of obstacles, and determine turning radius constraints, climb capability constraints, and safety interval constraints based on the aircraft performance parameters; a path generation module, used to generate candidate flight paths composed of multiple path nodes based on the environment model, and determine the local flyable space range at each path node; a recovery trajectory construction module, used to construct recovery trajectories for each path node, including at least one of a retreat trajectory, an ascent trajectory, and a lateral departure trajectory; a recovery capability assessment module, used to determine a set of feasible recovery trajectories for each path node based on the spatial connectivity of the recovery trajectories in the environment model, determine recovery difficulty parameters based on the spatial safety margin change trend of the recovery trajectories and the margin change characteristics that satisfy the turning radius constraints and climb capability constraints of the aircraft in the set of feasible recovery trajectories, and determine the path recoverability index of each path node in combination with the continuous reachability of the recovery trajectory to the preset safety area; and a path determination module, used to screen or optimize the candidate flight paths based on the path recoverability index of each path node to determine the target flight path. The recovery capability assessment module is specifically used to: discretize each recovery trajectory into multiple sampling points along the trajectory direction, and construct a local safety domain that satisfies the safety interval constraint at each sampling point; determine the spatial connectivity parameters of each recovery trajectory based on the intersection relationship between the local safety domain and the obstacle-occupied area in the environment model, including collision markers and continuous safe connectivity length; filter recovery trajectories according to the spatial connectivity parameters: when the collision marker indicates no collision and the continuous safe connectivity length is not less than a preset connectivity threshold, retain the corresponding recovery trajectory; and aggregate the retained recovery trajectories to obtain a set of feasible recovery trajectories corresponding to the path nodes. The recovery capability assessment module is further configured to: sample the recovery trajectories in the set of feasible recovery trajectories along the trajectory direction, calculate the spatial safety margin at each sampling point, and determine the minimum spatial safety margin and the rate of decrease of the safety margin as spatial safety margin change trend parameters based on the spatial safety margin sequence corresponding to each recovery trajectory; calculate the margin change parameters for the recovery trajectory to satisfy the turning radius constraint and the climbing capability constraint, and determine the recovery difficulty parameter corresponding to each recovery trajectory based on the spatial safety margin change trend parameter and the margin change parameter corresponding to the recovery trajectory; determine the trajectory recoverability value of each recovery trajectory based on the continuous reachable path length from the path node to the preset safe area and the corresponding recovery difficulty parameter, and select the maximum value of the trajectory recoverability value as the path recoverability index of the path node.

5. A low-altitude flight path planning platform according to claim 4, characterized in that: The recovery trajectory construction module is specifically used to: for any path node, extract the path segment corresponding to the preset backtracking distance along the reverse direction of the candidate flight path, and superimpose the braking distance determined by the flight speed and maximum deceleration at the path node into the backtracking distance to generate a backtracking trajectory; for any path node, construct an ascent segment from the path node along the altitude direction to a preset safe altitude, and construct an ascent safety body considering the safety interval constraint around the ascent segment, and generate an ascent trajectory based on the collision detection results of the ascent safety body and the environment model; for any path node, determine the lateral departure direction based on the heading direction of the path node, search for a target departure area that meets a preset openness threshold in the lateral departure direction, and generate a lateral departure trajectory connecting the path node and the target departure area under the condition of meeting the turning radius constraint.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 3.

Images