Low-altitude aircraft take-off and landing facility site selection method and device based on three-dimensional flight suitability evaluation

Abstract

The application discloses a low-altitude aircraft take-off and landing facility site selection method and device based on three-dimensional flight suitability evaluation, belongs to the technical field of low-altitude traffic system planning and low-altitude infrastructure site selection, and solves the technical problems that the volume, continuity and actual accessibility of a three-dimensional flyable airspace above a candidate facility are difficult to be quantified in the prior art. Multi-source spatial data is acquired, a candidate take-off and landing facility set is generated, a local three-dimensional voxel airspace model is constructed, hard non-passable marks and probability flyable attribute values are respectively generated based on three-dimensional airspace hard constraints and semantic constraints, coverage index, modified continuity index and accessibility index are calculated, a three-dimensional flyable airspace index is determined, five types of constraints, i.e., an unreachable rate, service coverage, construction quantity, safety distance and a minimum index threshold, are combined, candidate take-off and landing facilities are screened and layout optimized, and low-altitude aircraft take-off and landing facility site selection results are output. The application is used for realizing low-altitude aircraft take-off and landing facility site selection in a ground-airspace integrated manner.

Description

Technical Field

[0001] This invention relates to the field of low-altitude transportation system planning and low-altitude infrastructure site selection technology, specifically to a method and apparatus for site selection of low-altitude aircraft take-off and landing facilities based on three-dimensional flight suitability evaluation. Background Technology

[0002] With the development of demands related to low-altitude transportation, urban air traffic, and low-altitude logistics, low-altitude aircraft take-off and landing facilities have gradually become crucial infrastructure connecting ground transportation systems, urban functional spaces, and low-altitude flight activities. The site selection for low-altitude aircraft take-off and landing facilities involves not only the construction conditions of the ground or building infrastructure but also the accessibility, continuity, and actual service capacity of the three-dimensional airspace above and around the candidate location to meet demand units. Therefore, determining the location of low-altitude aircraft take-off and landing facilities within the complex urban built environment is a critical issue in low-altitude transportation infrastructure planning.

[0003] Currently, to address the site selection problem for low-altitude aircraft take-off and landing facilities, existing technologies primarily employ methods based on geographic information systems (GIS), demand distribution, and constraint optimization. For instance, Chinese patent document CN113407872A discloses a "Site Selection Method for Take-off and Landing Points of Urban Air Traffic Vehicles Based on POI," which screens and evaluates candidate take-off and landing points based on factors such as POI information, population distribution, land use type, and traffic conditions. This method mainly focuses on urban functions and ground suitability, enabling preliminary screening of candidate locations. However, its evaluation process remains concentrated in two-dimensional ground space, lacking a systematic analysis of the flyability of the three-dimensional airspace above the candidate locations.

[0004] For example, Chinese patent document CN120525580A discloses "a method for selecting take-off and landing points for UAVs in urban last-mile logistics". By constructing a customer buffer zone, a noise model and a safety constraint area, and optimizing the location results based on a grid map to consider cost and satisfaction, the method can take into account environmental constraints and service needs to a certain extent. However, its constraint expression is mainly processed uniformly in the form of a buffer zone, without distinguishing between constraints of different natures. Moreover, the location space is still mainly based on a two-dimensional grid, which makes it difficult to reflect the structural characteristics and channel continuity of the three-dimensional flight space.

[0005] In addition, Chinese patent document CN120235367A discloses "Intelligent Site Selection Method and System for UAV Take-off and Landing Points for Low-Altitude Logistics". It constructs a multi-dimensional decision model by introducing demand intensity, airspace status and environmental factors, and combines reinforcement learning or optimization algorithms to achieve site selection decision. It uses action masking to block invalid actions that violate airspace rules. In essence, it is a binary filtering mechanism, which is difficult to characterize continuous risk differences.

[0006] In summary, existing technologies have not yet formed a complete three-dimensional rational quantitative evaluation system for flyable airspace in complex urban environments, making it difficult to support the scientific site selection and dynamic optimization of low-altitude aircraft take-off and landing facilities in three-dimensional space.

[0007] Therefore, it is necessary to propose a site selection method for low-altitude aircraft take-off and landing facilities that can, based on initial ground screening, further conduct a joint evaluation of the size, continuity, and accessibility of the three-dimensional airspace above the candidate points, and has the ability to be implemented in an engineering manner and dynamically updated. Summary of the Invention

[0008] This invention solves the technical problem in the prior art that it is difficult to quantify the volume, continuity and actual accessibility of the three-dimensional flyable airspace above candidate facilities.

[0009] The method for selecting the location of low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation as described in this invention includes the following steps: Step 1: Acquire multi-source spatial data of the target area and generate a candidate set of take-off and landing facilities; Step 2: Determine the service evaluation domain for each candidate take-off and landing facility in the candidate take-off and landing facility set, and construct its corresponding local three-dimensional voxel spatial domain model. Step 3: Based on the local three-dimensional voxel airspace model, calculate the three-dimensional airspace index of the corresponding candidate take-off and landing facilities. Step 4: Based on the three-dimensional airspace index, select the final take-off and landing facility site selection scheme from the candidate take-off and landing facility set.

[0010] Furthermore, in one embodiment of the present invention, the step 1 of acquiring multi-source spatial data of the target area and generating a candidate take-off and landing facility set specifically includes: Based on multi-source spatial data of the target area, a pre-set hard constraint is set on the ground, and all take-off and landing positions that meet the pre-set hard constraint are selected. The selected take-off and landing positions constitute a candidate take-off and landing facility set.

[0011] Furthermore, in one embodiment of the present invention, the construction of the corresponding local three-dimensional voxel spatial model in step 2 specifically involves: The service evaluation domain is divided into the near-field core evaluation domain, the interface transition zone, and the far-field service evaluation domain. Based on the division results, a local three-dimensional voxel spatial domain model is constructed.

[0012] Furthermore, in one embodiment of the present invention, the local three-dimensional voxel spatial model is specifically: The system pre-defines three-dimensional airspace hard constraints and three-dimensional airspace semantic constraints. Based on the three-dimensional airspace hard constraints, it generates hard impassable identifiers for spatial units in the service evaluation domain. Based on the three-dimensional airspace semantic constraints, it calculates the obstacle membership degree of spatial units that are not eliminated by the hard impassable identifiers. Based on the obstacle membership degree, it determines the probabilistic flyability attribute value of the corresponding candidate take-off and landing facilities.

[0013] Furthermore, in one embodiment of the present invention, the calculation of the three-dimensional airspace index of the corresponding candidate take-off and landing facility in step 3 specifically includes: Based on the local three-dimensional voxel airspace model, the probabilistic flyability attribute value of the corresponding candidate take-off and landing facilities is calculated. Based on the probabilistic flyability attribute value, the coverage index, the modified continuity index, and the accessibility index are calculated. The coverage index, the modified continuity index, and the accessibility index are weighted and synthesized to obtain the three-dimensional flyability airspace index. The probabilistic flyable attribute value is specifically as follows: ; in, This is a probability-based flight attribute value. A set of semantic constraint levels. For the first Risk weight coefficients at the semantic constraint hierarchy level voxels Compared to the first Barrier membership degree of class semantic constraint hierarchy These are the index coordinates of the voxels in the 3D mesh.

[0014] Furthermore, in one embodiment of the present invention, step 4 involves selecting and optimizing the layout of candidate take-off and landing facilities from the candidate take-off and landing facility set, taking into account inaccessibility constraints, service coverage constraints, construction quantity upper limit constraints, safety distance constraints, and minimum index threshold constraints, to obtain the final take-off and landing facility site selection scheme. The minimum index threshold constraint is to remove candidate take-off and landing facilities whose three-dimensional airspace index is lower than the preset minimum index threshold.

[0015] Furthermore, in one embodiment of the present invention, the method further includes: Step 5: When a building change, flight restriction zone update or meteorological data refresh event is detected, identify candidate take-off and landing facilities within the scope of the event's impact, reuse the existing three-dimensional airspace index of unaffected candidate take-off and landing facilities, and recalculate the three-dimensional airspace index only for affected candidate take-off and landing facilities.

[0016] Furthermore, in one embodiment of the present invention, when there are multiple target aircraft types, the method further includes: For each target aircraft type, steps 1 to 3 are executed to obtain the three-dimensional airspace index corresponding to each aircraft type. For the same candidate take-off and landing facility, the three-dimensional airspace indexes obtained under multiple aircraft types are weighted and fused, or the minimum value is taken as the comprehensive three-dimensional airspace index of the candidate take-off and landing facility. The final take-off and landing facility site selection scheme in step 4 is based on the comprehensive three-dimensional airspace index of each candidate take-off and landing facility.

[0017] Furthermore, in one embodiment of the present invention, the coverage index is specifically: ; in, For coverage index, For the first Weight function for each height layer, voxels Clearance maneuverability Voxel volume; The modified continuity index is specifically: ; in, To correct the continuity index, This represents the fragmentation penalty intensity coefficient. This refers to the volume of the connected component where the candidate take-off and landing facilities are located. To effectively control the total amount of flight-grade follicles, The bottleneck accessibility of the main channel. For vertical continuity, This is an intermediate quantity of the fragmentation penalty factor; The accessibility index is specifically: ; in, For accessibility index, For candidate take-off and landing facilities up to the first The straight-line distance between the target points For reachability indicators, For candidate take-off and landing facilities up to the first The actual flight cost to each target point This represents the total number of target points generated within the target service area.

[0018] The low-altitude aircraft take-off and landing facility site selection device based on three-dimensional flightability evaluation described in this invention, the device for implementing the above method, includes the following modules: The ground-based preliminary screening module acquires multi-source spatial data of the target area and generates a set of candidate take-off and landing facilities. The local modeling module determines the service evaluation domain of each candidate take-off and landing facility in the candidate take-off and landing facility set, and constructs its corresponding local three-dimensional voxel spatial domain model. The index synthesis module calculates the three-dimensional airworthiness airspace index of the corresponding candidate take-off and landing facilities based on the local three-dimensional voxel airspace model. The layout optimization module selects the final take-off and landing facility site selection scheme from the candidate take-off and landing facility set based on the three-dimensional airspace index.

[0019] This invention solves the technical problem in existing technologies that make it difficult to quantify the volume, continuity, and actual accessibility of the three-dimensional flyable airspace above candidate facilities. Specific beneficial effects of this invention include: 1. This invention proposes a method for selecting low-altitude aircraft take-off and landing facilities based on three-dimensional airworthiness evaluation. By constructing a local three-dimensional voxel airspace model, calculating the three-dimensional airworthiness airspace index, and quantifying the volume, continuity, and actual accessibility of the three-dimensional flyable airspace above the candidate facilities, the traditional two-dimensional ground condition evaluation is extended into a comprehensive analysis framework of ground pre-constraints and local three-dimensional airspace evaluation, which can more realistically reflect the operational suitability of candidate take-off and landing facilities. 2. This invention proposes a site selection method for low-altitude aircraft take-off and landing facilities, introduces a probabilistic flyable attribute model and a semantic hierarchical obstacle model, and incorporates building entities, permanent regulatory restricted areas, conditional temporary restricted areas and meteorological constraints into a unified data interface, which solves the problems of insufficient expressive power of binary obstacle models and difficulty in unified coupling of different constraints in the prior art, and realizes adaptive updating of meteorological constraints through a dynamic time window statistical model. 3. This invention proposes a site selection method for low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation. Unlike existing technologies that rely solely on two-dimensional distance, construction cost, or simple coverage for site selection, this method utilizes coverage index CI and modified continuity index... The accessibility index AI is weighted and synthesized to obtain the three-dimensional airspace suitability index 3D. FI, where CI reflects the effective available space volume, The AI ​​reflects the continuity of effective channels, while the actual flight cost ratio is reflected. The three have a clear division of labor and a closed logic, which can scientifically compare and rank candidate take-off and landing facilities. 4. This invention proposes a site selection method for low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation. In the continuity evaluation, it introduces the proportion of effective connectivity, bottleneck accessibility, vertical continuity, and fragmentation penalty, which can identify narrow passages, vertical breaks, and isolated airspace problems caused by building obstruction or regulatory restrictions in high-density urban environments. In the accessibility evaluation, it introduces the actual flight cost composed of geometric length, probabilistic risk, insufficient clearance, vertical changes, and aircraft performance constraints, making the path evaluation closer to the actual low-altitude flight operation conditions, and supports adaptive loading of different aircraft parameters through aircraft configuration files. 5. This invention proposes a method for selecting low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation. Through an event-triggered dynamic update mechanism, only the affected voxels and the affected candidate take-off and landing facilities are locally recalculated. This method can achieve engineering-level maintenance and continuous updates in scenarios such as building changes, updates to flight restriction areas, and weather updates, and supports the long-term dynamic operation and management of low-altitude transportation infrastructure. Attached Figure Description

[0020] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart of the low-altitude aircraft take-off and landing facility site selection method based on three-dimensional flightability evaluation as described in Implementation Method 1. Figure 2 This is a schematic diagram of the three-zone division and semantic hierarchical voxelization described in Implementation Method 3, wherein (a) is a schematic diagram of the three-zone division and semantic hierarchical voxelization (top view), and (b) is a vertical cross-section of low-altitude airspace voxelization modeling and multi-source constraint layer. Figure 3 This is a schematic diagram of the event-triggered dynamic update hierarchical maintenance process described in Implementation Method Six; Figure 4 This is a schematic diagram of the address selection device module architecture described in Implementation Method 10. Detailed Implementation

[0021] Various embodiments of the present invention will now be clearly and completely described with reference to the accompanying drawings. The embodiments described with reference to the drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0022] Implementation Method 1: Currently, the existing technology mainly has the following problems: (1) Insufficient evaluation of three-dimensional airspace airworthiness: Existing methods are mostly based on two-dimensional space or simplified three-dimensional models for site selection analysis, which makes it difficult to uniformly quantify the three-dimensional airspace volume, spatial continuity and flyability quality on which low-altitude aircraft depend for operation, and cannot truly reflect the actual operating space of aircraft in complex environments. (2) Lack of a three-dimensional airspace quantification evaluation mechanism for site selection during the planning period: The site selection decision for low-altitude aircraft take-off and landing facilities occurs during the planning period, requiring the production of static, reproducible, cross-candidate horizontally comparable, and reportable item by item to the competent authorities three-dimensional airspace airspace airworthiness evaluation results for each candidate location. The output form of existing methods is structurally misaligned with this requirement: it is either a reinforcement learning strategy and action mask, or a multimodal dynamic temporal prediction, or a probability risk map on a two-dimensional grid. The above methods are either geared towards dynamic decision-making and situational awareness during the operation period, or generate evaluation results that are not easy to audit item by item through data-driven weighting mechanisms, and none of them provide joint quantitative indicators for the volume, channel continuity, and service accessibility of the three-dimensional airspace above the candidate location, making it difficult to directly support the selection, ranking, and layout optimization of candidate locations during the planning period; (3) Lack of 3D accessibility modeling: Existing methods mostly use points, grids or regions as basic units, lacking an integrated modeling approach oriented towards spatial units and path relationships, making it difficult to support 3D path search, channel analysis and spatial structure evaluation; (4) Service capability evaluation is still limited to two-dimensional level: Most methods are based on two-dimensional distance or service radius for accessibility analysis, without considering the real three-dimensional path formed by the aircraft when avoiding obstacles, avoiding restricted airspace and meeting flight performance constraints. This makes it difficult to accurately reflect the actual service accessibility and inaccessibility risk.

[0023] Therefore, to overcome the shortcomings of existing technologies that rely solely on two-dimensional ground conditions for site selection, neglect the three-dimensional airspace airworthiness above candidate points, lack unified quantitative evaluation indicators, and are difficult to dynamically maintain and update, this embodiment proposes a site selection method for low-altitude aircraft take-off and landing facilities based on a three-dimensional airworthiness airspace index, such as... Figure 1 As shown, this method is applicable to the planning, layout optimization, and dynamic updating of takeoff and landing facilities for UAVs, cargo aircraft, manned eVTOL (electric vertical takeoff and landing) aircraft, and other low-altitude aircraft. Through a local three-dimensional voxel airspace model surrounding candidate takeoff and landing facilities, and integrating constraints from buildings, flight restriction zones, weather conditions, and aircraft performance, this method quantitatively evaluates the effective availability, effective corridor continuity, and accessibility to the target service area of ​​the airspace above the candidate facilities. This supports the scientific site selection and continuous maintenance of urban low-altitude transportation infrastructure. Specifically, it includes the following: Step 1: Acquire multi-source spatial data of the target area and generate a candidate set of take-off and landing facilities; Step 2: Determine the service evaluation domain for each candidate take-off and landing facility in the candidate take-off and landing facility set, and construct its corresponding local three-dimensional voxel spatial domain model. For each candidate take-off and landing facility in the candidate take-off and landing facility set C, its service evaluation domain is determined, and a local three-dimensional voxel spatial domain model is constructed within the service evaluation domain; the local three-dimensional voxel spatial domain model includes multiple spatial units and path edges generated by adjacent spatial units; Step 3: Based on the local three-dimensional voxel airspace model, calculate the three-dimensional airspace index of the corresponding candidate take-off and landing facilities. Step 4: Based on the three-dimensional airspace index, select the final take-off and landing facility site selection scheme from the candidate take-off and landing facility set.

[0024] This implementation method can jointly and quantitatively evaluate the effective volume, channel continuity, and service accessibility of the three-dimensional flyable airspace above candidate take-off and landing facilities, thereby improving the scientific nature, comparability, and engineering applicability of take-off and landing facility site selection.

[0025] Implementation Method Two: The difference between this implementation method and Implementation Method One is that in step 1, multi-source spatial data of the target area is acquired to generate a candidate take-off and landing facility set, specifically as follows: Multi-source spatial data within the target area is acquired, construction carriers for supporting take-off and landing facilities are identified, candidate take-off and landing facility locations are generated and merged to extract representative points; a pre-constraint hard constraint screening of construction carriers is performed on the merged candidate take-off and landing facility locations to generate a candidate take-off and landing facility set. A demand unit set is generated based on traffic demand data and urban function data from the multi-source spatial data.

[0026] In this embodiment, the multi-source spatial data includes at least: building outline and height data, flight restriction zone boundary data, meteorological data, land use data, digital elevation model, power network node data, communication coverage data, and coordinate data of existing flight facilities.

[0027] The aforementioned pre-existing hard constraints on the ground include at least: 1. Land use suitability constraints; 2. Ground obstacle clearance constraints; 3. Infrastructure support constraints; 4. Safety distance constraints with existing flight facilities.

[0028] Potential takeoff and landing locations that do not meet any of the above constraints are directly eliminated; for locations that meet the constraints, a set of candidate takeoff and landing facilities is generated according to a preset sampling rule. : ; in, This represents the total number of candidate take-off and landing facilities.

[0029] In this embodiment, step 1 only deals with hard constraints that directly conflict with ground-based landing sites. Permanent regulatory restrictions, conditional temporary restrictions, and meteorological constraints in the airspace above candidate take-off and landing facilities are uniformly incorporated into the local three-dimensional voxel airspace model in subsequent steps.

[0030] Preferably, a hard constraint screening is performed on the ground first, and then a local three-dimensional voxel spatial model is constructed for the retained candidate take-off and landing facilities to reduce the computational overhead of full-area three-dimensional modeling, connectivity analysis and three-dimensional path planning, and improve the efficiency of site selection evaluation.

[0031] Implementation Method 3: The difference between this implementation method and Implementation Method 1 is that in step 2, the corresponding local three-dimensional voxel spatial model is constructed, specifically as follows: The service evaluation domain is divided into the near-field core evaluation domain, the interface transition zone, and the far-field service evaluation domain. Based on the division results, a local three-dimensional voxel spatial domain model is constructed.

[0032] The local three-dimensional voxel spatial model is specifically as follows: The system pre-defines three-dimensional airspace hard constraints and three-dimensional airspace semantic constraints. Based on the three-dimensional airspace hard constraints, it generates hard impassable identifiers for spatial units in the service evaluation domain. Based on the three-dimensional airspace semantic constraints, it calculates the obstacle membership degree of spatial units that are not eliminated by the hard impassable identifiers. Based on the obstacle membership degree, it determines the probabilistic flyability attribute value of the corresponding candidate take-off and landing facilities.

[0033] In this embodiment, for each candidate take-off and landing facility in the candidate take-off and landing facility set A local three-dimensional voxel spatial model is constructed above it within the horizontal and vertical ranges determined based on the target aircraft parameters and service area requirements.

[0034] To ensure both computational accuracy and efficiency, the service evaluation domain is divided into a near-field core evaluation domain. Interface transition zone And far-field service assessment domain Three areas, such as Figure 2 As shown, spatial units of different resolutions are used for modeling: Near-field core evaluation domain High-resolution modeling of the area surrounding candidate takeoff and landing facilities was performed using fine-grained voxel meshes. by Centered on, horizontal radius The cylindrical area uses a uniform side length Uniform fine voxel mesh modeling for coverage index The calculation, and as a corrected continuity index. Accessibility Index The basic fine mesh for far-field calculations; Interface transition band A portal node connects the near-field fine voxels and the far-field coarse nodes, and the portal node undertakes the semantic and geometric transformation between the near-field fine voxel set and the far-field coarse nodes: Horizontal radius range The annular region also uses unified fine voxel modeling for integration with the near-field core evaluation domain. Together they constitute the modified continuity index A unified fine voxel analysis domain is used to generate a set of interface portal nodes connecting near and far fields. ; Far-field service assessment domain Cross-domain spatial structure is represented by coarse nodes and the path edges between them, and the path edges are defined according to a uniform equivalent length metric: Horizontal radius range A ring-shaped region, with a side length not less than The bold voxels are used to abstractly represent the spatial domain, and then adjacent flyable bold voxels are further abstracted to generate a sparse coarse node graph, which is used for the accessibility index. Accessibility search in the far field segment.

[0035] In the near-field core evaluation domain Interface transition band Within this framework, semantic hierarchical voxels are used to classify building entities, permanent regulatory restriction zones, conditional temporary restriction zones, and meteorological constraints. The set of obstacle semantic levels is denoted as […]. At least including: Building solid layer, permanent regulatory restriction layer, conditional temporary restriction layer, and weather constraint layer. For any voxel Define its probability flyable attribute value for: ; in, For the first Risk weight coefficients at the semantic level. voxels Compared to the first Barrier membership degree at the semantic level of a class.

[0036] Preferably, the local three-dimensional voxel spatial model uses three-dimensional voxels as the basic spatial units to represent the local spatial domain in a rasterized manner; a higher resolution is used in areas with dense obstacles, boundary areas of flight restriction zones, and areas with drastic changes in spatial constraints, while a lower resolution is used in areas with smaller changes in spatial constraints, in order to improve computational accuracy and efficiency. Each voxel records the actual voxel side length. and voxel volume .

[0037] Therefore, the flyability property of voxels is extended from traditional binary logic to... The continuous probability value within the interval is used to uniformly support subsequent evaluations of coverage, continuity, and accessibility. The probabilistic attribute value... Together with voxel space coordinates, they constitute a unified input interface for the three-dimensional voxel spatial domain model.

[0038] For the meteorological constraint layer, the obstacle membership degree is preferably determined using a dynamic time window statistical model: ; in, voxels exist Wind field intensity at any given moment This refers to the rated maximum permissible wind field threshold for the corresponding aircraft model. To preset a historical time window. During the static site selection phase... During the dynamic operation and update phase, it can be expanded to... .

[0039] By As a static input to the meteorological constraint layer, the probabilistic flyability attribute value is substituted. The computational formula, together with other semantic constraint layers, constitutes a unified input interface for the three-dimensional voxel spatial domain model.

[0040] for The bold elements in the text, if their average probability of flight is not less than the navigation threshold, are considered valid. Average net airspace maneuverability not less than And the equivalent channel width is not less than Then, coarse nodes are generated by modifying the center of the coarse voxel. If the axis-aligned bounding boxes of two coarse nodes corresponding to coarse voxels have a non-zero volume intersection or the center distance does not exceed a preset threshold, and the space traversed by the line connecting the two centers satisfies the navigable condition, then a graph edge is established between them. Its equivalent length is uniformly defined as: ; Generation of the interface portal node: in the interface transition zone Extract those that meet the navigation threshold Flyable fine voxel collection For flyable fine voxel sets Perform 3D connected component analysis; for each axis-aligned bounding box connected to the near-field principal connected component and to a coarse node in the far field, there is a non-zero volume intersection or the center distance does not exceed a preset adjacency threshold. Connected subsets Generate a portal node Its representative coordinates are taken from a connected subset. The probability-weighted centroid of the inner voxel center can be determined by the property: ; in, For the first Interface portal node The three-dimensional representation coordinates, Belongs to a connected subset A specific voxel in the, voxels The probability of flying attribute value, voxels volume, voxels The center three-dimensional coordinates.

[0041] Each portal node Simultaneously record the average probability of flyable attribute value. Average headroom availability Altitude range Representative channel width and the set of far-field coarse nodes in contact with it. Only when Only then can the portal node participate in subsequent cross-domain path calculations to ensure that the portal itself meets the requirements for navigation probability, clearance, and channel width.

[0042] The spatial coordinates, volume, and probabilistic flyability of each voxel Together, they constitute a local three-dimensional voxel spatial model corresponding to the candidate takeoff and landing facilities. This model is both the coverage index of the three sub-indicators and the coverage index of the candidate facilities. Corrected continuity index and accessibility index It serves as a common input and a unified data interface for event-triggered dynamic updates.

[0043] Implementation Method Four: The difference between this implementation method and Implementation Method One is that, in step 3, the three-dimensional airworthiness airspace index for the corresponding candidate take-off and landing facility is calculated as follows: Based on the local three-dimensional voxel airspace model, the probabilistic flyability attribute values ​​of the corresponding candidate take-off and landing facilities are calculated. Based on the probabilistic flyability attribute values, the coverage index, the modified continuity index, and the accessibility index are calculated. The coverage index, the modified continuity index, and the accessibility index are weighted and synthesized to obtain the three-dimensional flyability airspace index.

[0044] In this embodiment, the coverage index, the corrected continuity index, and the accessibility index are specifically as follows: 1. Coverage Index

[0045] Coverage Index This is used to characterize the effective available airspace volume around candidate takeoff and landing facilities. For each voxel in the local three-dimensional voxel airspace model, a coverage index is defined by weighting it together with the probabilistic flyability attribute value, voxel volume, altitude layer weight, and clearance maneuverability. for: ; in, For coverage index, voxels Clearance maneuverability Voxel volume For the first The weighting function for each altitude level is preferably a Gaussian weighting form centered on the typical flight altitude of the aircraft. ; in, For the first The flight altitude corresponding to the upper level The typical flight altitude of the target aircraft type. denoted as the standard deviation of the height distribution; K represents the total number of height layers.

[0046] voxels The clearance maneuverability is preferably defined as: ; in, The distance from the voxel center to the nearest obstacle voxel. This refers to the minimum safe clearance distance for the corresponding aircraft model.

[0047] The coverage index It not only reflects the size of the flyable airspace around candidate take-off and landing facilities, but also takes into account altitude preference, obstacle clearance margin and multi-resolution voxel volume difference, thus it can better characterize the effective available airspace size than simple probability accumulation or binary voxel counting.

[0048] 2. Corrected continuity index

[0049] Modified continuity index Used to characterize the continuity of effective passageways around candidate take-off and landing facilities, The computation is performed on a uniform fine voxel mesh to eliminate adjacency ambiguity caused by directly performing connectivity analysis on multi-resolution voxels. Based on a preset threshold... By filtering the probabilistic flyability attributes, a set of effective flyable voxels is obtained. : ; Then, a connected component analysis is performed on the set of effective flyable voxels according to the three-dimensional adjacency relationship to obtain the total number of effective flyable voxels: ; And the effective volume of the connected component where the candidate take-off and landing facilities are located: ; in, For the first One connected component; To further characterize the impact of narrow passages, vertical breaks, and fragmentation on flight continuity, the main passage bottleneck navigability is defined as follows: ; in, This represents the minimum equivalent passage width of the main passage corresponding to the connected component where the candidate take-off and landing facility is located. This refers to the minimum maneuvering aisle width required for the corresponding aircraft type.

[0050] Preferably, the The results are obtained by performing cross-sectional scanning, skeleton extraction, or local minimum net width estimation on the main passageway between the candidate take-off and landing facility and the target service direction within the connected component where the candidate take-off and landing facility is located.

[0051] Define vertical continuity as: ; in, The number of altitude layers required to maintain connectivity with the connected component containing the candidate takeoff and landing facilities. To assess the total number of height levels; Define the intermediate value of fragmentation penalty as: ; in, This refers to the number of small connected components whose voxel count or voxel volume is less than a preset threshold. The total number of connected components; Then correct the continuity index Determine using the following formula: ; in, To correct the continuity index, This represents the fragmentation penalty intensity coefficient. This refers to the volume of the connected component where the candidate take-off and landing facilities are located. To effectively control the total amount of flight-grade follicles, The bottleneck accessibility of the main channel. For vertical continuity, This is an intermediate quantity of the fragmentation penalty factor; because Calculated on a uniform fine-grained mesh, volumetric statistics are based on... As a unit, Through statistical analysis of cross-sections layer by layer along the centerline of the main channel, and using... The unit of measurement is unique, the calculation definition is unique, and the results achieved independently by different engineers are numerically consistent.

[0052] The modified continuity index It no longer focuses solely on topological judgments of connectivity, but comprehensively characterizes the main connectivity volume, channel bottleneck characteristics, vertical continuity characteristics, and fragmentation degree of the effective flyable airspace surrounding candidate take-off and landing facilities, thus better reflecting the effective channel continuity required for the actual operation of low-altitude aircraft.

[0053] 3. Accessibility Index

[0054] Accessibility Index To characterize the actual flight cost ratio of candidate takeoff and landing facilities to the target service area, a hierarchical path planning strategy combining near-field fine voxels and far-field coarse node graphs is adopted, through the interface portal node set. By connecting the two ends of the path, the problem of inconsistent costs for cross-scale movement can be fundamentally eliminated.

[0055] With the candidate take-off and landing facility as the center, the radius The target service area is defined as the region, and Poisson disk sampling is used to generate the sample. There are 1 target point, and the flight altitude for each target point is uniformly set to... .

[0056] For the Target points :like lie in Within this framework, 3D A* path planning is directly performed on a uniform, fine-grained mesh; if lie in Inside, the near-field segment is planned on a uniform fine voxel grid extending to the near-field exit voxel. Then, based on location and height similarity, they are mapped to the optimal portal node. The far-field segment from Starting from the corresponding far-field coarse node, search on the sparse coarse node graph to... Far-field node If there is no far-field coarse node that meets the conditions, it is directly marked as unreachable.

[0057] No. The complete actual flight cost to each target point is: ; in, For near-field flight costs, For the cost of cross-domain transition flights, Cost of far-field segment flight.

[0058] For any path edge Its equivalent length is uniformly defined as the actual Euclidean distance between the two representative points: ; in, and They are nodes and nodes The three-dimensional spatial coordinates.

[0059] The step size is not inherited from either side to eliminate the influence of the fine voxel step size: Generate a set of target points within the target service area, denoted as _____. A set of target points. Preferably, the set of target points is generated through uniform sampling or Poisson disk sampling, and the flight altitude of the target points is consistent with the flight altitude of typical missions.

[0060] Starting from candidate takeoff and landing facilities, three-dimensional path planning is performed within the local three-dimensional voxel airspace model to solve for feasible flight paths to each target point. Preferably, the three-dimensional path planning uses a 26-neighborhood extended three-dimensional... Algorithm, 3D Algorithms or other equivalent 3D obstacle avoidance planning algorithms; during path planning, voxel flyability attributes, clearance availability, vertical variation, and aircraft performance constraints all participate in path cost calculation. The above formula is equivalent to accumulating each cost component along the path edge by edge; in this implementation, the definition is given in the form of component summation.

[0061] For the For each target point, the actual flight cost is expressed as: ; in, For the path geometric length components, For the probabilistic risk equivalent length component. For the equivalent length component of insufficient clearance, The equivalent length component for vertical variation. Equivalent length component constrained by model performance; The above equivalent length component is preferably defined as follows: ; ; ; ; in, From candidate take-off and landing facilities to the first The three-dimensional path to the target point For the path through voxels The equivalent step size, For the path in voxels Vertical change at point The function is the penalty function for performance constraints of the aircraft model; Accordingly, the accessibility index for: ; in, For accessibility index, For candidate take-off and landing facilities up to the first The straight-line distance between the target points As an reachability indicator, when the target point is reachable. Unreachable time , For candidate take-off and landing facilities up to the first The actual flight cost to each target point This represents the total number of target points generated within the target service area.

[0062] The accessibility index It is composed of the ratio of straight-line distance to actual flight cost, where the numerator is the ideal direct distance and the denominator is the equivalent flight cost after considering geometric detours, probabilistic risks, insufficient airspace, vertical changes and aircraft performance constraints. Therefore, compared with the traditional ratio of straight-line distance to actual path length, it can better reflect the operational accessibility of candidate take-off and landing facilities in real three-dimensional restricted airspace.

[0063] Coverage Index Corrected continuity index and accessibility index Weighted synthesis was performed to obtain the three-dimensional airspace index of candidate take-off and landing facilities. : ; in, And satisfy: ; Preferably, when the number of candidate take-off and landing facilities is not less than the preset minimum sample size, the weighting coefficient is... The entropy weight method is used to adaptively determine the weights based on the evaluation matrix of candidate take-off and landing facilities; when the sample size is insufficient or the application scenario has clear business preferences, preset weights can also be used.

[0064] Through the above synthesis method, the Candidate take-off and landing facilities are uniformly scored based on three dimensions: effective available airspace volume, effective channel continuity, and actual flight cost ratio.

[0065] Implementation Method 5: The difference between this implementation method and Implementation Method 1 is that in step 4, the candidate take-off and landing facilities are screened and their layout optimized by combining inaccessibility rate constraints, service coverage constraints, construction quantity limit constraints, safety distance constraints, and minimum index threshold constraints to obtain the final take-off and landing facility site selection scheme. The minimum index threshold constraint is to remove candidate take-off and landing facilities whose three-dimensional airspace index is lower than the preset minimum index threshold.

[0066] In this embodiment, after obtaining each candidate take-off and landing facility... Subsequently, by combining constraints such as inaccessibility rate, service coverage, maximum number of facilities to be built, safety distance, and minimum index threshold, candidate take-off and landing facilities are screened and their layout optimized. Preferably, the following optimization problem can be constructed: ; satisfy ; in, For the collection of candidate take-off and landing facilities, This is the final set of candidate take-off and landing facilities. To set an upper limit for the planned construction quantity, For the minimum service coverage target, The minimum safe distance between take-off and landing facilities. The inaccessibility rate of candidate take-off and landing facility c. This is the upper limit threshold for the unreachability rate. This is the minimum threshold for the three-dimensional airspace suitability index.

[0067] Through the above constraint optimization, the final take-off and landing facility site selection scheme can be obtained while meeting the requirements of service coverage and operational safety.

[0068] Implementation Method Six: The difference between this implementation method and Implementation Method One is that the method further includes: Step 5: When a building change, flight restriction zone update or meteorological data refresh event is detected, identify candidate take-off and landing facilities within the scope of the event's impact, reuse the existing three-dimensional airspace index of unaffected candidate take-off and landing facilities, and recalculate the three-dimensional airspace index only for affected candidate take-off and landing facilities.

[0069] Most existing methods are based on static data and perform one-time calculations, which are difficult to cope with dynamic factors such as weather changes, temporary airspace restrictions or changes in demand, and lack the ability to quickly update local changes and reuse results.

[0070] Therefore, in order to solve the above-mentioned technical problems, in this embodiment, when a building change, flight restriction zone update, or meteorological data refresh event is detected, the affected voxel set is identified. Only the obstacle membership degree and probabilistic flyability attribute of the semantic constraint layer are recalculated for the affected voxel set, and the existing index results and path planning cache results of unaffected candidate take-off and landing facilities are reused. Only the coverage index of the affected candidate take-off and landing facilities is recalculated. Corrected continuity index and accessibility index and three-dimensional airspace index ,like Figure 3 As shown.

[0071] Through the above partial update strategy, this implementation method is not only applicable to one-time site selection assessment, but also to the continuous maintenance and dynamic adjustment after the low-altitude transportation infrastructure is put into operation.

[0072] In this embodiment, the site selection method is a static planning approach at the decision output level, and its flightability indicators are obtained by aggregating the dynamic operating environment through historical joint statistical windows. The membership degree of the meteorological constraint layer obstacle is determined by... Time points earned: ; in, For historical wind field observation time window, voxels Wind field intensity at any given moment This refers to the rated maximum permissible wind field threshold for the corresponding model.

[0073] After statistical integration, its output does not depend on time variables. It participates in the site selection calculation through a unified interface together with the static membership of the building entity layer, the permanent regulatory restriction layer, and the conditional temporary restriction layer. When the building changes, the restricted area is updated, or the meteorological data is refreshed, the affected candidate take-off and landing facilities are partially recalculated by the event triggering mechanism. This recalculation does not change the static attributes of the site selection decision.

[0074] Implementation Method Seven: The difference between this implementation method and Implementation Method One is that when there are multiple target aircraft types, the method further includes: For each target aircraft type, steps 1 to 3 are executed to obtain the three-dimensional airspace index corresponding to each aircraft type. For the same candidate take-off and landing facility, the three-dimensional airspace indexes obtained under multiple aircraft types are weighted and fused, or the minimum value is taken as the comprehensive three-dimensional airspace index of the candidate take-off and landing facility. The final take-off and landing facility site selection scheme in step 4 is based on the comprehensive three-dimensional airspace index of each candidate take-off and landing facility.

[0075] Aircraft model-related parameters should include at least: typical flight altitude Standard deviation of height distribution Maximum permissible wind field threshold Minimum safe clearance distance Minimum maneuvering lane width requirement Vertical climb or descent capability constraints and accessibility index Cost weighting coefficients for each equivalent length component , , , The above parameters are adaptively loaded based on the device model configuration file to support the differentiated evaluation needs of different device models.

[0076] In summary, this invention, based on a local three-dimensional voxel spatial domain model of candidate takeoff and landing facilities, innovatively divides the local spatial domain into a three-zone structure: a near-field core evaluation domain, an interface transition zone, and a far-field service evaluation domain. A unified, uniform, fine-grained voxel mesh is constructed in the near field to ensure computational accuracy; an interface portal node set is generated in the transition zone to eliminate cross-scale adjacency ambiguity; and a sparse, coarse-grained node graph is used in the far field to improve the efficiency of large-scale path planning. Furthermore, using a probabilistic flyable attribute model and a semantic hierarchical obstacle model as unified input interfaces, it analyzes coverage index (CI) and modified continuity index... The candidate take-off and landing facilities are quantitatively evaluated using three dimensions: accessibility index, AI, and three-dimensional airspace suitability index. The system integrates FI synthesis, takeoff and landing facility screening and layout optimization, and event-triggered dynamic updates. It also employs a dual expression of hard inaccessibility constraints (voxel removal) and continuous risk constraints to characterize the difference between rigid inaccessibility and continuous risk, thereby enabling the expansion from ground-based site selection to integrated ground-airspace site selection.

[0077] Implementation Method 8: A site selection device for low-altitude aircraft take-off and landing facilities, the device being used to implement the method described in Implementation Method 1, comprising the following modules: The ground-based preliminary screening module is used to acquire multi-source spatial data within the study area and perform pre-construction hard constraint screening on potential take-off and landing locations to generate a set of candidate take-off and landing facilities. The local modeling module is used to construct local three-dimensional voxel airspace models for each candidate take-off and landing facility. It performs semantic hierarchical voxelization of building entities, permanent regulatory restriction zones, conditional temporary restriction zones, and meteorological constraints, and assigns probabilistic flyable attribute values ​​to each voxel. The index calculation module is used to calculate the coverage index based on the local three-dimensional voxel spatial domain model. Corrected continuity index and accessibility index ; The index synthesis module is used to synthesize coverage indices. Corrected continuity index and accessibility index Weighted synthesis is performed to obtain the three-dimensional airspace index for flight suitability. ; The layout optimization module is used to optimize the layout based on the three-dimensional airspace index. By combining inaccessibility constraints, service coverage constraints, safety distance constraints, and minimum index threshold constraints, candidate take-off and landing facilities are screened and their layout optimized, and a take-off and landing facility site selection scheme is output. Furthermore, in one embodiment of this implementation, the apparatus further includes: The dynamic update module is used to identify the affected voxel set when it detects changes in buildings, updates to flight restriction areas, or refreshes in meteorological data. It then locally recalculates the various indices of the relevant candidate take-off and landing facilities and the three-dimensional airspace index to update the take-off and landing facility site selection plan.

[0078] Implementation Method Nine: This implementation method is a specific embodiment of the above-mentioned method for selecting the location of low-altitude aircraft take-off and landing facilities.

[0079] Example 1: Site selection for urban low-altitude aircraft take-off and landing facilities based on three-dimensional airspace suitability index This embodiment takes the core area of ​​a city and its surrounding expansion area as the research object, and the research scope is approximately... The target aircraft is a multi-rotor cargo low-altitude aircraft. The main parameters preset in its configuration file are: typical flight altitude. Standard deviation of height distribution Minimum safe clearance distance Minimum low-altitude airspace width requirement Maximum permissible wind field threshold Maximum permissible climb or descent angle .

[0080] The multi-dimensional spatial data used in this embodiment includes: building outline and height data, permanent regulatory restriction zone boundary data, conditional temporary restriction zone announcement data, historical wind field data for the past year and predicted wind field data for the next 48 hours, urban land use data, digital elevation model, power grid node and remaining capacity data, 4G / 5G communication coverage data, and coordinate data of existing flight facilities. All of the above data are uniformly projected onto the CGCS2000 plane coordinate system and calibrated to a unified elevation datum.

[0081] Step 1: Construction carrier identification, candidate facility generation, and pre-existing hard constraint screening This step is used to identify construction carriers capable of supporting low-altitude aircraft take-off and landing facilities from the study area, generate candidate take-off and landing facility locations, and perform pre-construction hard constraint screening on them.

[0082] Step 1.1: Construction Carrier Identification Based on building 3D contour and height data, land use data, planned reserved construction land data, and digital elevation models, construction carriers suitable for supporting low-altitude aircraft take-off and landing facilities are identified. These construction carriers include: building roofs that meet structural load-bearing requirements; planned reserved land for low-altitude take-off and landing facilities; land for commercial and service facilities; industrial land; transportation hub land; land for public management and public service facilities; and other construction spaces permitted by planning regulations for the installation of low-altitude take-off and landing facilities.

[0083] In this embodiment, approximately 1,500 potential take-off and landing facility locations were identified and extracted.

[0084] Step 1.2: Generation of candidate take-off and landing facility locations and selection of representative points For each construction site, one or more candidate take-off and landing facility locations are generated based on its geometric center, available roof area, optimal clearance location, power access direction, and main approach and departure directions.

[0085] For multiple candidate locations located within the same construction site, or locations whose horizontal distance to each other is less than a preset proximity distance threshold, a merging process is performed. In this embodiment, the preset proximity distance threshold is set to 100 meters. During merging, candidate locations that meet the following conditions are preferentially selected as representative points: fewer obstacles within the ground protection area; greater distance from surrounding tall buildings; better power and communication access conditions; and better spatial connectivity with the main service direction.

[0086] After merging and extracting representative points, 980 candidate take-off and landing facility representative locations were obtained.

[0087] Step 1.3: Screening of Pre-construction Hard Constraints for Construction Carriers First, land suitability constraints. If a candidate site is located within highly sensitive land such as residential land, ecological protection land, cultural relic protection land, hazardous chemical storage and utilization land, school land, or hospital land, it will be eliminated; if it is located within commercial service facilities land, industrial land, transportation facilities land, public service facilities land, or comprehensive land permitted by planning, it will be retained.

[0088] Second, ground obstacle clearance constraints. A ground takeoff and landing support zone with a radius of 50 meters is established centered on the candidate location. Fixed obstacles higher than 5 meters and located in the main approach and departure directions are identified within the support zone. If such obstacles exist, the candidate location is eliminated.

[0089] Third, infrastructure support constraints. Within a 200-meter radius of the candidate location, search for distribution boxes, substations, or other accessible power nodes to determine if their available reserved capacity is no less than 50kW. Simultaneously, deploy 20 sampling points within a 300-meter communication support radius to determine if the RSRP (Reference Signal Received Power) is no less than [value missing]. The proportion of sampling points with a strength of 105dBm is not less than 90%. If either power access or communication coverage fails to meet this requirement, the candidate location is eliminated.

[0090] Fourth, safety distance constraints with existing flight facilities. Calculate the horizontal distance between the candidate location and existing helicopter landing pads, known low-altitude takeoff and landing facilities, and other flight facilities. If the horizontal distance is less than 800 meters, the candidate location is eliminated.

[0091] After the above-mentioned pre-construction hard constraint screening, this embodiment finally obtains the candidate take-off and landing facility set: After the above initial screening, this embodiment retains a total of The candidate take-off and landing facilities are denoted as: ; Step 1.4: Generation of Demand Unit Set Based on traffic demand data and urban function data from multi-source spatial data, the study area is divided into multiple demand units. These demand units can be determined according to population density, logistics demand intensity, distribution of commercial service facilities, distribution of transportation hubs, distribution of industrial parks, and distribution of public service facilities.

[0092] In this embodiment, the study area is discretized into a set of demand units: ; Each demand unit Assigning weights to demand It is used for subsequent service coverage calculation and layout optimization.

[0093] Step 2: Determining the service evaluation domain and constructing a local 3D voxel spatial model For candidate take-off and landing facility set Each candidate take-off and landing facility The service evaluation domain is determined, and a local three-dimensional voxel spatial model is constructed within the service evaluation domain.

[0094] In this embodiment, candidate take-off and landing facilities are used. Centered on a central point, a service evaluation domain with a horizontal radius of 3000 meters and a vertical range extending from the ground to 120 meters was constructed. This service evaluation domain was divided into a near-field core evaluation domain, an interface transition zone, and a far-field service evaluation domain, as shown in Table 1. Table 1

[0095] Step 3: Calculation of hard impassable markers, obstacle membership, and probability of flightability. This step is used to generate hard impassable identifiers for spatial cells based on three-dimensional spatial hard constraints, and to calculate the obstacle membership degree and probabilistic flyability attribute values ​​of spatial cells based on three-dimensional spatial semantic constraints.

[0096] Step 3.1: Identification of hard constraints in three-dimensional spatial domain For any spatial cell v, a hard impassable identifier H(v) = 1 is generated if it satisfies any of the following conditions: The spatial unit is located inside the building entity; the spatial unit is located inside a permanently prohibited flight area; the height of the spatial unit is lower than the ground elevation corresponding to the digital elevation model; the path edge between the spatial unit and the adjacent spatial unit crosses a hard impassable area; the climb angle or descent angle corresponding to the path edge between the spatial units exceeds the allowable threshold of the target aircraft type.

[0097] If the above conditions are not met, then H(v) = 0. When H(v) = 1, the spatial cell is removed or the generation of path edges associated with it is prohibited.

[0098] Step 3.2, Three-dimensional spatial semantic constraint layer For spatial units not removed by the hard impassable identifier, the barrier membership degree is further calculated based on the semantic constraint layer. The set of semantic constraint layers is denoted as: ; Among them, s=1 is the building entity layer, s=2 is the permanent regulatory restriction layer, s=3 is the conditional temporary restriction layer, and s=4 is the meteorological constraint layer.

[0099] For any spatial unit v(i,j,k), its probabilistic flyability attribute value is defined as: ; in, Let v be the obstacle membership degree of spatial unit v relative to the s-th semantic constraint layer.

[0100] For a building's solid floor, if the spatial unit overlaps with the building's solid floor, then If the spatial unit does not enter the building entity but is within the building boundary buffer distance; Within a certain range, the rate decreases linearly based on the distance to the building boundary; if it exceeds the buffer distance, then... .

[0101] For permanent regulatory restrictions, if the spatial unit is located inside the permanent regulatory restriction zone, then If the space unit is located outside the buffer zone boundary of the permanent regulatory restriction area. Within a range of meters, the decay rate decreases linearly with distance; if it exceeds the buffer distance, then... .

[0102] For conditional temporary constraint layers, both spatial impact and temporal validity are considered. Let the overlap duration between the temporary constraint event and the evaluation time window be denoted as . The total length of the evaluation time window is The event validity coefficient is: ; set up Let v be the spatial membership degree of spatial unit v relative to the temporary restricted area and its boundary buffer range, then: ; The preferred buffer distance for the temporary restricted area boundary is 100 meters.

[0103] For the meteorological constraint layer, a historical time window statistical model is used: ); in, Let be the wind field intensity of spatial unit v at time t. For historical observation time window, The rated maximum permissible wind field threshold for the corresponding model. In this embodiment, Historical wind field data from the past year is used. When this method is applied to the dynamic operation and updating of existing facilities, ,in, The future forecast time window can be obtained from the wind field data predicted for the next 48 hours.

[0104] In this way, hard constraints are used to eliminate spatial units or prohibit the generation of path edges, while semantic constraints are used to generate continuous obstacle membership degrees and probabilistic flyable attribute values, thus together forming a unified input interface for the three-dimensional airspace navigable model.

[0105] Step 4: Generation of near-field and far-field nodes, path edges, and interface portal nodes Step 4.1, Generation of far-field coarse node graph Within the far-field service assessment domain, the average probabilistic flyability attribute, average clearance maneuverability, and equivalent aisle width are calculated for each bold voxel. A far-field coarse node is generated if the bold voxel simultaneously satisfies the following conditions: ; ; ; in, The average probability of a bold element being able to fly is a property. The bold element represents the average net airspace maneuverability. This is the equivalent channel width.

[0106] If the bounding boxes of the coarse voxels corresponding to the two far-field coarse nodes have a non-zero volume intersection, or the distance between the centers of the two coarse voxels does not exceed 50 meters, and the space traversed by the line connecting the two satisfies the navigable condition, then a path edge is established between the two coarse nodes.

[0107] The equivalent length of any path edge e=(u,v) is uniformly defined as the three-dimensional Euclidean distance between the two representative points: ; Step 4.2, Generation of Interface Portal Nodes In the interface transition zone, extract the set of flyable fine voxels that meet the navigation threshold: ; right Perform three-dimensional 26-neighborhood connected component analysis. For each connected subset... An interface portal node will be generated if it meets all of the following conditions. : Connected to the near-field main connectivity region; The axis-aligned bounding boxes of at least one far-field coarse node have a non-zero volume intersection with the corresponding coarse voxel, or the center distance does not exceed a preset adjacency threshold. The average probability of flightability, average clearance maneuverability, and representative channel width all meet the navigation threshold.

[0108] Interface portal node The representative coordinates are taken from the center of the fine voxel within a connected subset, which is a probability that can be expressed by the weighted centroid of the property: ; Each interface portal node records the following attributes: average probability of flyability; average clearance maneuverability; altitude band range; representative channel width; and set of adjacent far-field coarse nodes.

[0109] Only when an interface portal node simultaneously meets the probabilistic flyability threshold, the airspace maneuverability availability threshold, and the channel width threshold can it participate in subsequent cross-domain path calculations.

[0110] Step 5: Calculation of Coverage Index, Adjusted Continuity Index, and Accessibility Index For each candidate take-off and landing facility The coverage index is calculated based on its three-dimensional spatial accessibility model. Corrected continuity index and accessibility index .

[0111] Step 5.1, Coverage Index The coverage index is used to characterize the volume of effective available airspace around candidate takeoff and landing facilities. In this embodiment, the coverage index is calculated on a uniform fine voxel grid in the near-field core evaluation domain.

[0112] The near-field core evaluation domain is vertically divided into K=24 height layers, each with a height of 5 meters. The height weight of the k-th height layer is: ; in, Let the center height of the k-th height layer be _____. The typical flight altitude of the target aircraft is used in this embodiment. = 80 m; For the standard deviation of the height distribution, this embodiment takes... = 30 m; Voxel clearance maneuverability is defined as: ; in, This is the distance from the center of a spatial unit to the nearest obstacle spatial unit or obstacle boundary. rice.

[0113] The coverage index is calculated as follows: ; in, Let V(v) be the set of voxels within the k-th height layer, and let V(v) be the volume of the spatial unit v.

[0114] Step 5.2, Correct the continuity index The modified continuity index is used to characterize the continuity of effective pathways around candidate takeoff and landing facilities. In this embodiment, it is calculated on a uniform fine voxel grid in the near-field core evaluation domain and the interface transition zone to avoid adjacency ambiguity caused by direct connectivity analysis of multi-resolution voxels.

[0115] First, the effective set of flyable voxels is obtained based on the probabilistic flyability attribute threshold: ; in, .

[0116] right Connectivity component analysis was performed based on the 26-neighborhood adjacency relationships in three dimensions, resulting in several connected components: ; Define the volume percentage of the connected component where the candidate take-off and landing facility is located as follows: ; in, For the first The number or volume of voxels in a connected component.

[0117] Bottleneck mobility is defined as: ; in, Connectivity component where candidate take-off and landing facilities are located The minimum equivalent passage width of the inner main passageway corresponding to the main passage. rice, Meters. The aforementioned It can be obtained through main passageway cross-section scanning, skeleton extraction, or local minimum net width estimation.

[0118] Define vertical continuity as: ; in, The number of altitude layers required to maintain connectivity with the connected component containing the candidate takeoff and landing facilities. To assess the total number of height levels.

[0119] The fragmentation penalty factor is defined as: ; in, This refers to the number of small connected components whose voxel count or voxel volume is less than a preset threshold. The total number of connected components. .

[0120] The corrected continuity index is then: ; This index not only determines whether airspace is connected, but also characterizes the main connectivity volume, channel bottlenecks, vertical breaks, and degree of fragmentation.

[0121] Step 5.3 Accessibility Index The accessibility index is used to characterize the ratio of the actual flight cost of a candidate takeoff and landing facility to the target service area.

[0122] This embodiment focuses on candidate takeoff and landing facilities. Generate M=150 target points within the target service area: ; The target point can be generated using Poisson disk sampling, and the height of the target point is set to 80 meters.

[0123] For the j-th target point The actual flight cost is calculated using a hierarchical path planning method.

[0124] If the target point is located within the near-field core evaluation domain or the interface transition zone, then 3D A* path planning is performed directly on the uniform fine voxel mesh.

[0125] If the target point is located within the far-field service evaluation domain, it is first mapped to a far-field coarse node that meets the height tolerance, navigation threshold, and clearance threshold. Then, in the near-field fine voxel grid, candidate take-off and landing facilities are planned to near-field exit voxels; next, it is mapped to the optimal interface portal node. Finally, starting from the far-field coarse node corresponding to the interface portal node, a search is performed on the sparse coarse node graph to find the far-field coarse node corresponding to the target point. If no far-field coarse node or interface portal node meets the conditions, the target point is recorded as unreachable.

[0126] To avoid confusion with variable subscripts, this embodiment uses candidate take-off and landing facilities. To the target point The optimal portal node is denoted as: ; in, A set of optional portal nodes that meet connectivity, height tolerance, probabilistic flyability thresholds, clearance thresholds, and channel width thresholds; The height difference penalty coefficient; This represents the height of the portal node. The height of the target point. If... If empty, then the target point For candidate facilities Unreachable.

[0127] For any path edge e=(u,v), its equivalent length is: ; The path edge cost is defined as: ; ; in, The average probability of flyability of the spatial units at both ends of the path edge. The average clearance maneuverability of the space units at both ends of the path edge. The height difference between adjacent nodes The function is the penalty function for performance constraints of the aircraft model. This is the cost weighting coefficient.

[0128] Candidate take-off and landing facilities To the target point The actual flight cost is: ; in, For the near-field segment, The cost of cross-domain portal mapping This represents the cost of the far-field segment. The cost of each segment is obtained by summing the costs of the corresponding path edges: ; Candidate take-off and landing facilities To the target point The three-dimensional straight-line distance is denoted as: ; Define reachable identifier When a feasible 3D path exists that satisfies the constraints, When the route is unreachable or the actual flight cost exceeds the preset limit, .

[0129] The accessibility index is then: ; And define candidate take-off and landing facilities The unreachability rate is: ; Step 6: Synthesis of three-dimensional airspace suitability index For each candidate take-off and landing facility Coverage Index Corrected continuity index and accessibility index The evaluation feature vector is composed of these features.

[0130] When the number of candidate takeoff and landing facilities is not less than the preset minimum sample size, the entropy weight method is used to determine the index weights. Specifically, an evaluation matrix is ​​constructed: ; in, For the first The candidate takeoff and landing facilities were in the first Scores on each evaluation metric is the row number of the matrix.

[0131] ; Normalizing the evaluation matrix yields: ; Calculate the first Information entropy of each indicator: ; in, The normalization constant is For the first The candidate facilities in the Normalized weighting of each indicator; ; in, For the first The candidate facilities in the Normalized scores on each indicator; Calculate the coefficient of difference: ; And calculate the indicator weights: ; in, For the first The coefficient of variation of each evaluation indicator.

[0132] ; in, For the first The final weighting coefficients of each evaluation indicator.

[0133] When the number of candidate take-off and landing facilities is less than the preset minimum sample size, or when the planning scenario has clear business preferences, preset weights can also be used.

[0134] Ultimately, candidate take-off and landing facilities The three-dimensional airspace suitability index is: ; in, The larger the value, the better the three-dimensional airspace flightability of the candidate take-off and landing facility.

[0135] Step 7: Screening and Layout Optimization of Takeoff and Landing Facilities After obtaining the three-dimensional airspace index of each candidate take-off and landing facility, the candidate take-off and landing facilities are screened and their layout optimized by combining inaccessibility constraints, service coverage constraints, safety distance constraints and minimum index threshold constraints.

[0136] Define decision variables When candidate take-off and landing facilities When selected, ,otherwise Define the requirement unit coverage variables. When demand unit When covered by the final solution, ,otherwise .

[0137] The following optimization objective is proposed: ; in, This is a balance coefficient between service coverage targets and airworthiness index targets.

[0138] The constraints include: First, service coverage constraints: ; In this embodiment, .

[0139] Second, the upper limit constraint on the number of construction projects: ; In this embodiment, .

[0140] Third, minimum safety distance constraints between takeoff and landing facilities: ; In this embodiment, rice.

[0141] Fourth, the unreachability constraint: ; In this embodiment, .

[0142] Fifth, minimum airworthiness index threshold constraint: ; In this embodiment, .

[0143] This embodiment uses the branch and bound method to solve the above optimization model and obtain the final landing facility site selection scheme. It should be noted that the 100-meter merging threshold in step 1 is used to extract representative points of the same construction carrier or adjacent candidate locations, and the 500-meter safety distance constraint in step 7 is used for operational safety isolation in the final layout scheme. The two are constraints at different levels.

[0144] Step 8: Event-triggered dynamic update When events such as building changes, flight restriction zone updates, or weather data refreshes are detected, event-triggered dynamic updates are performed.

[0145] First, identify the set of affected voxels based on the scope of the event's impact: ; If candidate take-off and landing facilities Service evaluation domain and If the candidate take-off and landing facilities do not intersect, the existing index results and path planning cache results of the candidate take-off and landing facilities are reused; if they intersect, the candidate take-off and landing facilities are marked as affected candidate take-off and landing facilities.

[0146] For the affected candidate take-off and landing facilities, only the obstacle membership degree and probabilistic flyability attribute values ​​of the semantic constraint layer are recalculated for the affected voxel set: ; If changes within the near-field core evaluation domain lead to a change in the boundary of the near-field main connectivity domain, then the set of interface portal nodes for the affected candidate take-off and landing facilities will be regenerated.

[0147] If the voxels within the interface transition zone change, only the affected portal nodes are updated, including recalculating the representative coordinates, average probability of flyability, average clearance maneuverability, altitude band range, representative channel width, and set of adjacent far-field coarse nodes for the portal nodes.

[0148] If voxels within the far-field service evaluation domain change, local reconstruction is performed only on the affected coarse nodes and affected coarse graph edges.

[0149] For Non-intersecting historical paths are cached and reused directly; for paths that pass through affected voxels, affected portal nodes, or affected coarse graph edges, path planning is re-executed only for the relevant target points.

[0150] Subsequently, the coverage index, corrected continuity index, accessibility index, and three-dimensional airspace index were recalculated only for the affected candidate take-off and landing facilities, and the necessary layout optimization or partial replacement was re-executed based on the updated index results.

[0151] Through the aforementioned event-triggered dynamic update mechanism, this embodiment can achieve local recalculation and cache reuse in scenarios such as building changes, adjustments to permanent regulatory restricted areas, the activation or deactivation of conditional temporary restriction events, and meteorological data updates, thereby improving the efficiency of dynamic maintenance of take-off and landing facility site selection schemes.

[0152] This embodiment enables the screening of ground conditions for low-altitude aircraft take-off and landing facilities, three-dimensional airspace accessibility modeling, evaluation of effective available airspace volume, evaluation of effective channel continuity, evaluation of service accessibility, synthesis of three-dimensional airspace suitability index, layout optimization, and event-triggered dynamic updates in complex urban built environments.

[0153] Compared with methods that rely solely on ground conditions or two-dimensional service radius for site selection, this embodiment comprehensively considers construction carrier conditions, building entities, permanent regulatory restrictions, conditional temporary restrictions, meteorological constraints, airspace conditions, channel continuity, aircraft performance constraints, and actual three-dimensional path costs, thereby improving the safety, comparability, engineering applicability, and dynamic maintenance capabilities of low-altitude aircraft take-off and landing facility site selection.

[0154] Implementation Method 10: This implementation method is a specific embodiment of the above-mentioned low-altitude aircraft take-off and landing facility site selection device.

[0155] Example 2: Low-altitude aircraft take-off and landing facility site selection device based on three-dimensional flight suitability index This embodiment provides a site selection device for low-altitude aircraft take-off and landing facilities based on a three-dimensional airspace suitability index, used to implement the site selection method described in Embodiment 1. Figure 4As shown, the device includes a ground-based pre-screening module, a local modeling module, an index calculation module, an index synthesis module, a layout optimization module, and a dynamic update module.

[0156] The ground-based preliminary screening module is used to acquire multi-source spatial data within the study area, identify construction carriers for supporting low-altitude aircraft take-off and landing facilities based on the multi-source spatial data, generate candidate take-off and landing facility locations based on the construction carriers, and merge and extract representative points from candidate take-off and landing facility locations within the same construction carrier or within a preset proximity range; perform pre-construction hard constraint screening on the merged candidate take-off and landing facility locations, and output a set of candidate take-off and landing facilities. Simultaneously, a set of demand units is generated based on traffic demand data and urban function data. .

[0157] The pre-construction hard constraint screening includes land suitability constraints, ground obstacle clearance constraints, infrastructure support constraints, and safety distance constraints from existing flight facilities. If any constraint is not met, the corresponding candidate location will be directly eliminated.

[0158] The input to the ground-based preliminary screening module is multi-source spatial data, and the output is a set of candidate take-off and landing facilities. and demand unit set .

[0159] The local modeling module is used to receive the set of candidate take-off and landing facilities output by the ground pre-screening module, determine the service evaluation domain for each candidate take-off and landing facility, and construct a three-dimensional airspace navigability model within the service evaluation domain.

[0160] The three-dimensional spatial traversable model includes multiple spatial units and path edges generated from adjacent spatial units. The service evaluation domain is divided into a three-region structure: the near-field core evaluation domain, the interface transition zone, and the far-field service evaluation domain. The near-field core evaluation domain and the interface transition zone are modeled using uniform thin voxels with uniform side lengths, while the far-field service evaluation domain is modeled using coarse voxels and generates a sparse coarse node graph.

[0161] The local modeling module further includes the following sub-units: The hard constraint identifier sub-unit is used to generate hard impassable identifiers H(v) for spatial units based on three-dimensional spatial hard constraints, and to remove or prohibit the generation of path edges for spatial units that meet the hard impassable conditions. The semantic constraint and probabilistic attribute calculation subunit is used to calculate the obstacle membership degree of each spatial unit based on four types of semantic constraint layers: building entity layer, permanent regulatory restriction layer, conditional temporary restriction layer, and meteorological constraint layer. and probability of flying attribute value ; The node and path edge generation sub-unit is used to generate far-field coarse nodes and coarse node graph edges that meet the navigation threshold in the far-field service evaluation domain. In the interface transition zone, it generates an interface portal node set based on three-dimensional connected component analysis. Each portal node records the average probability of flyability attribute value, average airspace maneuverability availability, altitude band range, representative channel width, and adjacent far-field coarse node set.

[0162] The input to the local modeling module is a set of candidate take-off and landing facilities and multi-source spatial data. The output is a three-dimensional airspace navigable model corresponding to each candidate take-off and landing facility, including the spatial coordinates, voxel volume, probabilistic flyable attribute value, path edge set, and interface portal node set of each spatial unit.

[0163] The index calculation module receives the three-dimensional airspace accessibility model output by the local modeling module, and calculates the coverage index, modified continuity index, and accessibility index for each candidate take-off and landing facility based on this model. The index calculation module includes the following sub-units: The coverage index calculation sub-unit is used to calculate the coverage index on a uniform fine voxel grid in the near-field core evaluation domain, using probabilistic flyability attribute values, voxel volume, altitude layer weights, and clearance maneuverability availability as weights. ; The modified continuity index calculation subunit is used to obtain the modified continuity index on a unified fine voxel grid of the near-field core evaluation domain and the interface transition zone through effective flyable voxel screening, three-dimensional 26-neighborhood connectivity component analysis, effective connectivity volume ratio calculation, bottleneck mobility calculation, vertical continuity calculation, and fragmentation penalty factor calculation. ; The accessibility index calculation subunit is used to generate a set of target points within the target service area, calculate the actual flight cost from candidate take-off and landing facilities to each target point using a hierarchical path planning method, and calculate the accessibility index based on the ratio of the actual flight cost to the three-dimensional straight-line distance. Simultaneously output the inaccessibility rate of each candidate take-off and landing facility. .

[0164] The input to the index calculation module is a three-dimensional airspace accessibility model, and the output is the coverage index, modified continuity index, accessibility index, and inaccessibility rate for each candidate take-off and landing facility.

[0165] The index synthesis module receives three indices output by the index calculation module, and performs weighted synthesis of the coverage index, the corrected continuity index, and the accessibility index to obtain a three-dimensional airspace index for each candidate takeoff and landing facility: ; in, , When the number of candidate take-off and landing facilities is not less than the preset minimum sample size, the weighting coefficient is adaptively determined based on the candidate take-off and landing facility evaluation matrix using the entropy weighting method; when the sample size is insufficient or the planning scenario has clear business preferences, the preset weights are used.

[0166] The index synthesis module takes three index values ​​for each candidate takeoff and landing facility as input and outputs a three-dimensional airspace suitability index for each candidate takeoff and landing facility. .

[0167] The layout optimization module is used to receive the three-dimensional airspace index output by the index synthesis module and the inaccessibility rate output by the index calculation module, and combine it with the demand unit set output by the ground front screening module to screen and optimize the layout of candidate take-off and landing facilities.

[0168] The layout optimization module constructs and solves an optimization problem with the following constraints: service coverage constraint, upper limit constraint on the number of facilities to be built, minimum safe distance constraint between take-off and landing facilities, inaccessibility constraint and minimum airworthiness index threshold constraint, and outputs the final take-off and landing facility site selection scheme.

[0169] The layout optimization module takes as input a set of candidate take-off and landing facilities and their three-dimensional airspace index, inaccessibility rate, and demand unit set, and outputs the final selected set of take-off and landing facilities and their spatial layout scheme.

[0170] The dynamic update module is used to identify the set of affected voxels based on the scope of the event when it detects events such as building changes, updates to flight restriction zones, or refreshed meteorological data. .

[0171] The dynamic update module performs the following hierarchical update operations: If the service evaluation domain of a candidate take-off and landing facility does not intersect with the set of affected voxels, the existing index results and path cache results of the candidate take-off and landing facility are reused; if they intersect, the candidate take-off and landing facility is marked as an affected candidate take-off and landing facility, and only the semantic constraint layer obstacle membership degree and probabilistic flyable attribute value are recalculated for the set of affected voxels; if changes in the near-field core evaluation domain cause changes in the boundary of the near-field main connected domain, the interface portal node set of the affected candidate take-off and landing facility is regenerated; if voxels in the interface transition zone or far-field service evaluation domain change, only the graph edges of the affected portal nodes and affected coarse nodes are reconstructed.

[0172] The dynamic update module then calls the index calculation module and the index synthesis module to recalculate the coverage index, corrected continuity index, accessibility index and three-dimensional airspace index of the affected candidate take-off and landing facilities, and calls the layout optimization module to perform local layout replacement or global re-optimization when necessary.

[0173] The dynamic update module takes event change information and the set of affected voxels as input and outputs the updated take-off and landing facility site selection scheme.

[0174] The data flow between the above six modules is as follows: ground-level preliminary screening module → local modeling module → index calculation module → index synthesis module → layout optimization module. The dynamic update module interacts with the local modeling module, index calculation module, index synthesis module, and layout optimization module when an event is triggered. Each module can be implemented in software, dedicated hardware, or a combination of software and hardware.

[0175] All or part of the functions of the various methods in the above embodiments can be implemented by hardware or by computer programs. When all or part of the functions in the above embodiments are implemented by computer programs, the program can be stored in a computer-readable storage medium, and the program can be executed by a computer to achieve the above functions. The computer program includes one or more computer instructions, which can be executed on the following processors: general-purpose processors, special-purpose processors, single-processor systems, multi-processor systems, distributed computing systems, or any combination of the above systems.

[0176] In addition, when all or part of the functions in the above embodiments are implemented by computer programs, the programs can also be stored in storage media such as servers, other computers, disks, optical discs, flash drives, or portable hard drives. They can be downloaded or copied to the memory of the local device, or the system of the local device can be updated. When the program in the memory is executed by the processor, all or part of the functions in the above embodiments can be implemented.

[0177] The above provides a detailed description of the method and apparatus for selecting the location of low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation proposed in this invention. Specific examples have been used to illustrate the principles and implementation methods of this invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. A method for site selection of low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation, characterized in that, Includes the following steps: Step 1: Acquire multi-source spatial data of the target area and generate a candidate set of take-off and landing facilities; Step 2: Determine the service evaluation domain for each candidate take-off and landing facility in the candidate take-off and landing facility set, and construct its corresponding local three-dimensional voxel spatial domain model. Step 3: Based on the local three-dimensional voxel airspace model, calculate the three-dimensional airworthiness airspace index of the corresponding candidate take-off and landing facilities, specifically as follows: Based on the local three-dimensional voxel airspace model, the probabilistic flyability attribute value of the corresponding candidate take-off and landing facilities is calculated. Based on the probabilistic flyability attribute value, the coverage index, the modified continuity index, and the accessibility index are calculated. The coverage index, the modified continuity index, and the accessibility index are weighted and synthesized to obtain the three-dimensional flyability airspace index. The probabilistic flyable attribute value is specifically as follows: ； in, This is the probability of being able to fly attribute value. It is a set of semantic constraint levels. For the first Risk weight coefficients at the semantic constraint hierarchy level voxels Compared to the first Barrier membership degree of class semantic constraint hierarchy These are the index coordinates of the voxels in the 3D mesh; Step 4: Based on the three-dimensional airspace index, select the final take-off and landing facility site selection scheme from the candidate take-off and landing facility set.

2. The method for site selection of low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation according to claim 1, characterized in that, Step 1 involves acquiring multi-source spatial data of the target area and generating a candidate set of takeoff and landing facilities, specifically as follows: Based on multi-source spatial data of the target area, a pre-set hard constraint is set on the ground, and all take-off and landing positions that meet the pre-set hard constraint are selected. The selected take-off and landing positions constitute a candidate take-off and landing facility set.

3. The method for site selection of low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation according to claim 1, characterized in that, Step 2 involves constructing the corresponding local three-dimensional voxel spatial model, specifically as follows: The service evaluation domain is divided into the near-field core evaluation domain, the interface transition zone, and the far-field service evaluation domain. Based on the division results, a local three-dimensional voxel spatial domain model is constructed.

4. The method for selecting the location of low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation according to claim 1 or 3, characterized in that, The local three-dimensional voxel spatial model is specifically as follows: The system pre-defines three-dimensional airspace hard constraints and three-dimensional airspace semantic constraints. Based on the three-dimensional airspace hard constraints, it generates hard impassable identifiers for spatial units in the service evaluation domain. Based on the three-dimensional airspace semantic constraints, it calculates the obstacle membership degree of spatial units that are not eliminated by the hard impassable identifiers. Based on the obstacle membership degree, it determines the probabilistic flyability attribute value of the corresponding candidate take-off and landing facilities.

5. The method for site selection of low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation according to claim 1, characterized in that, Step 4 involves selecting and optimizing the layout of candidate take-off and landing facilities from the candidate set, taking into account inaccessibility constraints, service coverage constraints, construction quantity limits, safety distance constraints, and minimum index threshold constraints, to obtain the final take-off and landing facility site selection scheme. The minimum index threshold constraint is to remove candidate take-off and landing facilities whose three-dimensional airspace index is lower than the preset minimum index threshold.

6. The method for selecting the location of low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation according to claim 1, characterized in that, The method further includes: Step 5: When a building change, flight restriction zone update or meteorological data refresh event is detected, identify candidate take-off and landing facilities within the scope of the event's impact, reuse the existing three-dimensional airspace index of unaffected candidate take-off and landing facilities, and recalculate the three-dimensional airspace index only for affected candidate take-off and landing facilities.

7. The method for site selection of low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation according to claim 1, characterized in that, When there are multiple target aircraft types, the method further includes: For each target aircraft type, steps 1 to 3 are executed to obtain the three-dimensional airspace index corresponding to each aircraft type. For the same candidate take-off and landing facility, the three-dimensional airspace indexes obtained under multiple aircraft types are weighted and fused, or the minimum value is taken as the comprehensive three-dimensional airspace index of the candidate take-off and landing facility. The final take-off and landing facility site selection scheme in step 4 is based on the comprehensive three-dimensional airspace index of each candidate take-off and landing facility.

8. The method for site selection of low-altitude aircraft take-off and landing facilities based on three-dimensional flightability evaluation according to claim 1, characterized in that, The coverage index is specifically: ； in, For coverage index, For the first Weight function for each height layer, voxels Clearance maneuverability Voxel volume; The modified continuity index is specifically: ； in, To correct the continuity index, This represents the fragmentation penalty intensity coefficient. This refers to the volume of the connected component where the candidate take-off and landing facilities are located. To ensure the total amount of effective flyable proteins, The bottleneck accessibility of the main channel. For vertical continuity, This is an intermediate quantity of the fragmentation penalty factor; The accessibility index is specifically: ； in, For accessibility index, For candidate take-off and landing facilities up to the first The straight-line distance between the target points For reachability indicators, For candidate take-off and landing facilities up to the first The actual flight cost to each target point This represents the total number of target points generated within the target service area.

9. A site selection device for low-altitude aircraft take-off and landing facilities based on three-dimensional flightworthiness evaluation, the device being used to implement the method described in claim 1, characterized in that, Includes the following modules: The ground-based preliminary screening module acquires multi-source spatial data of the target area and generates a set of candidate take-off and landing facilities. The local modeling module determines the service evaluation domain of each candidate take-off and landing facility in the candidate take-off and landing facility set, and constructs its corresponding local three-dimensional voxel spatial domain model. The index synthesis module calculates the three-dimensional airworthiness airspace index of the corresponding candidate take-off and landing facilities based on the local three-dimensional voxel airspace model. The layout optimization module selects the final take-off and landing facility site selection scheme from the candidate take-off and landing facility set based on the three-dimensional airspace index.

Images