A three-dimensional space grid-based urban low-altitude falling risk assessment method

Abstract

This invention discloses a method for assessing urban low-altitude fall risks based on a three-dimensional spatial grid, belonging to the field of low-altitude flight risk assessment technology. It aims to address the problem of accurately assessing low-altitude fall risks in complex urban scenarios. The invention includes: 1) 3) Rasterizing the low-altitude airspace of the assessment area into voxels, obtaining a voxel set; 2) quantifying population density grids or mobile signaling-derived population, sensitive facility vectors, tree canopy shading, and building shelters formed by three-dimensional building models, and assigning these values ​​to the voxels of the assessment airspace, resulting in a voxel-based environmental parameter database covering the entire area; 3) constructing an aircraft failure rate model; 4) conducting an uncertainties simulation experiment of the aircraft's fall trajectory to obtain the probability density distribution of the aircraft's landing coverage and terminal velocity; 5) considering the probability of collision with people, synthesizing voxel risks, and classifying the voxel risks to complete an urban low-altitude fall risk assessment based on a three-dimensional spatial grid.

Description

Technical Field

[0001] This invention belongs to the field of low-altitude flight risk assessment technology, specifically relating to a method for assessing urban low-altitude fall risks based on a three-dimensional spatial grid. Background Technology

[0002] Low-altitude crash risk assessment is one of the key issues in the safe management of urban air traffic operations. In typical ultra-high-density urban environments, with highly concentrated populations and numerous buildings, aircraft system failures or power outages can easily cause significant damage to ground personnel and infrastructure. Therefore, scientifically quantifying crash risk is a core prerequisite for the large-scale development of the low-altitude economy.

[0003] In international regulatory systems, both the International Civil Aviation Organization (ICAO) and the European Union Aviation Safety Agency (EASA) consider crash risk as a crucial factor in the approval, route planning, and operational restrictions for unmanned aerial vehicles (UAVs) and urban air traffic. The Specific Operations Risk Assessment (SORA) framework has been widely adopted for systematically assessing the operational risks of UAVs under different operating environments and mission conditions. The SORA method primarily categorizes risks based on factors such as the operating environment, aircraft characteristics, and risk mitigation measures, presenting the results in a qualitative or semi-quantitative classification matrix. While this method is mature and capable of meeting the needs of macro-level management and operational approval, its underlying principles remain largely unchanged.

[0004] While the Specific Operational Risk Assessment (SORA) framework has become a widely adopted tool for drone operational risk management internationally, this method has the following shortcomings in complex urban scenarios:

[0005] 1. Primarily qualitative, lacking refined quantitative analysis. The core of SORA is to classify operational activities into low, medium, and high risk categories through a hierarchical matrix. Its assessment results are mainly qualitative or semi-quantitative conclusions, lacking specific spatial quantitative indicators, which makes it difficult to meet the needs of refined management in high-density operating environments.

[0006] 2. Two-dimensional perspective, lacking three-dimensional spatial representation. The ground risk level (GRC) calculation in the SORA framework is mainly based on two-dimensional factors such as the area and population density of the operating area, failing to establish a unified three-dimensional spatial modeling system, and therefore unable to characterize the risk differences at different altitude levels. For megacities with high population and building density, the two-dimensional perspective cannot fully reflect the risk distribution characteristics of low-altitude operations.

[0007] 3. Static assessment, lacking dynamic adaptability. SORA typically relies on fixed data and scenario assumptions before operation, and the assessment results remain unchanged throughout the operation, making it difficult to reflect the dynamic impact of population distribution and environmental conditions on risk levels over time.

[0008] Therefore, the SORA framework is suitable for providing risk classification reference in macro-management and operational approval. However, its qualitative, two-dimensional and static deficiencies make it difficult to meet the refined risk assessment needs required for the normalization and large-scale operation of the low-altitude economy in the future. Summary of the Invention

[0009] The problem this invention aims to solve is the accurate assessment of low-altitude fall risk in complex urban scenarios, and proposes a method for assessing urban low-altitude fall risk based on a three-dimensional spatial grid.

[0010] To achieve the above objectives, the present invention provides the following technical solution:

[0011] A method for assessing urban low-altitude fall risk based on a three-dimensional spatial grid includes the following steps:

[0012] S1. The low-altitude airspace of the evaluation airspace is rasterized into three-dimensional voxels of the evaluation airspace to obtain the voxel set of the evaluation airspace.

[0013] S2. Quantify the population inversion population, sensitive facility vector, tree canopy shading, and building shelter formed by the three-dimensional building model from the population density grid or mobile signaling, and assign the values ​​to the voxels of the evaluation spatial domain obtained in step S1 to obtain a voxelized environmental parameter database covering the entire domain.

[0014] S3. Construct an aircraft failure rate model;

[0015] S4. Based on the aircraft failure altitude obtained in step S3, conduct an uncertainty simulation test of the aircraft's fall trajectory to obtain the aircraft's landing coverage probability density distribution and terminal velocity.

[0016] S5. Combine the voxelized environmental parameter database covering the entire domain obtained in step S2 with the aircraft failure rate obtained in step S3 and the aircraft crash trajectory obtained in step S4, and consider the probability of collision with people to perform voxel risk synthesis to obtain voxel risk.

[0017] S6. Perform risk classification on the voxels obtained in step S5 to complete an urban low-altitude fall risk assessment based on a three-dimensional spatial grid.

[0018] Furthermore, the specific implementation method of step S1 includes the following steps:

[0019] S1.1. Select the evaluation airspace The altitude range is from 100 m to 600 m above the ground, covering the typical operating altitudes of mainstream drone logistics, urban air traffic (UAM), and inspection aircraft.

[0020] Set voxel side length ,in Let x be the x-axis side length of the voxel. Let be the y-axis side length of the voxel. Let z be the z-axis side length of the voxel. The side length;

[0021] S1.2. Divide the evaluation spatial domain into sets of voxels at equal intervals. ,in, For voxels, for Any one of them, To determine the number of voxels that divide the study region on the x-axis, for Any one of them, To determine the number of voxels that divide the study region along the y-axis, for Any one of them, The number of voxels that divide the study area along the z-axis;

[0022] Set the voxel center coordinates to ,in Here are the coordinates of the starting reference point on the x-axis for the study area. Let the coordinates of the starting reference point on the y-axis be the coordinates of the study area. The coordinates of the starting reference point of the study area on the z-axis;

[0023] Set row-major one-dimensional encoding , It is a ternary index;

[0024] S1.3. Perform a voxel validity assessment. If a voxel intersects with a building entity, a no-fly zone, or a terrain obstacle, it is marked as invalid; otherwise, it is valid. At the same time, based on management requirements, extend the building horizontally outward by 10-20 m to form a safety buffer zone to obtain valid voxels.

[0025] S1.4. Spatial indexing and topological association are performed using 6-adjacent-face adjacency based on effective voxels.

[0026] Furthermore, the specific implementation method of step S2 includes the following steps:

[0027] S2.1. Perform data alignment and rasterization to obtain the corresponding mobile signaling inversion population. Sensitive facility vector Three-dimensional architectural models form architectural shelters ;

[0028] S2.2. Computational Fusion and Exposure, expressed as:

[0029]

[0030]

[0031] in, For integration and Exposure Population inversion for interval-normalized mobile signaling, For interval-normalized sensitive facility vectors, The weights for the population derived from inter-regional normalized mobile signaling. The weights of the interval-normalized sensitive facility vector;

[0032] S2.3. Collect the building shelter coefficient, which is obtained by empirical regression of building volume, height, and materials. ;

[0033] Calculate tree canopy shading coefficient: Calculate Sentinel-2 spectral coefficient NDVI to estimate canopy coverage. The tree canopy shading coefficient is obtained by comparing the mean NDVI with the mean NDVI. :

[0034]

[0035] in, The mean of the Sentinel-2 spectral coefficients. These are empirical parameters;

[0036] Calculate the shading factor The resulting expression is:

[0037] ;

[0038] S2.4. Voxelize the exposure and shading coefficients to obtain the voxelized exposure. and the occlusion coefficient after voxel assignment The expression is:

[0039]

[0040]

[0041] in, This represents the average exposure level.

[0042] S2.5. Divide the spatial domain corresponding to the raster into cubic voxel grids, and integrate all the information obtained in steps S2.1 to S2.5 in the raster to obtain a voxelized environmental parameter database covering the entire domain.

[0043] Furthermore, the specific implementation method of step S3 includes the following steps:

[0044] S3.1. Collect the aircraft's model mean time between failures (MTBF) and failure mode set. The failure modes include power interruption, loss of control, and energy depletion. (Task profile) Determine the set of failure modes The failure mode weights are derived from the aircraft type airworthiness data and historical operational records to obtain the base failure rate of each failure mode. ;

[0045] S3.2. Constructing the aircraft failure rate model yields:

[0046]

[0047]

[0048] in, For aircraft failure rate, Let m be the weight of the m-th failure mode. Let m be the base failure rate for the m-th failure mode. The task profile weight function, It represents the center height of the k-th height layer.

[0049] Furthermore, the simulation input for step S4 is the wind field profile parameters and gust intensity. Based on the relevant duration T, aerodynamic parameter range, initial attitude / velocity range, building / terrain voxels and effective voxel domain, the system is set to terminate upon ground contact and records end-effector velocity samples. and layer influence area The hit count is obtained by convolving the landing kernel of each sample with the ground grid and then normalizing it into a landing coverage probability density distribution. .

[0050] Furthermore, the specific implementation method of step S5 includes the following steps:

[0051] S5.1. Calculate the probability of colliding with a person, using the following expression:

[0052]

[0053] in, is the probability of collision with people, is the population density, is the layer influence area;

[0054] S5.2. Calculate the casualty probability, and the expression is:

[0055]

[0056]

[0057]

[0058] Among them, is the casualty probability, is the kinetic energy at the end of the layer; are the basic fatality rate parameter, energy sensitivity parameter, and shelter effect parameter respectively; is the mass of the drone, is the S-shaped logistic function;

[0059] S5.3. Perform voxel risk synthesis to obtain the risk of the voxel , and the expression is:

[0060] .

[0061] Furthermore, the specific implementation method of step S6 is to set the first threshold as T1, the second threshold as T2, and the third threshold as T3, where the threshold T1 < T2 < T3, and perform risk grading on the risk of the voxel obtained in step S5 to obtain:

[0062]

[0063] Among them, is the risk level of the voxel.

[0064] Advantages of the present invention:

[0065] The method for evaluating the risk of low-altitude ground impact in cities based on three-dimensional spatial grids described in the present invention solves the problem of the lack of refined quantitative indicators in the prior art; the present invention outputs an absolute risk value of 0-1 on a unified three-dimensional voxel and supports multi-scale aggregation of columns / layers / regions, forming indicators such as "total layer risk, regional mean / quantile, hot spot list" that can be directly used for approval and evaluation. For any candidate route, the route risk cost is accumulated according to the sequence of traversed voxels, which can be directly docked with the threshold (TLS) or the "red-yellow-green" rule, supporting batch comparison and rapid decision-making. Under the voxel configuration of 50 m × 50 m × 50 m, the spatial positioning error of the high-risk area does not exceed one voxel; when needed, it can be switched to 25 m voxels to achieve block-level accuracy.

[0066] This invention presents a three-dimensional spatial grid-based method for assessing urban low-altitude fall risks, addressing the shortcomings of existing two-dimensional representations and the difficulty in depicting height differences. Using height layers as organizational units, this invention assesses impact point coverage and personnel exposure layer by layer, constructing a three-dimensional risk volume and layer-slice maps to intuitively reflect the risk differences and vertical gradients of the same plane location at different height layers. It can identify high-risk / low-risk height zones, providing quantitative basis for layered flight path layout, takeoff and landing procedure height restrictions, and layered airspace access. By uniformly modeling building shelter / shading and crowd exposure within column domains and consistently assigning values ​​to each height layer, it can significantly reduce misjudgments caused by "plane averaging" in densely built-up areas.

[0067] This invention presents a method for assessing urban low-altitude fall risks based on a three-dimensional spatial grid, addressing the problems of existing assessments being static and difficult to update with changing scenarios. This invention uses day / night or time-of-day population layers and wind field scenarios as external parameters, supports time-slice recalculation and outputs time-varying risk maps, adapting to pre-operation assessments, in-operation reassessments, and emergency reassessments. It supports comparisons across multiple scenarios such as "normal / strong winds" and "weekdays / weekends," quantifying risk fluctuation ranges and providing evidence for differentiated flow control / detours / windowed releases. Attached Figure Description

[0068] Figure 1 This is a flowchart of a method for assessing the risk of low-altitude falls to the ground in cities based on a three-dimensional spatial grid, as described in this invention. Detailed Implementation

[0069] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention; that is, the described specific embodiments are merely a part of the embodiments of the invention, and not all of them. The components of the specific embodiments of the invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations, and the invention may also have other embodiments.

[0070] Therefore, the following detailed description of specific embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected specific embodiments of the invention. All other specific embodiments obtained by those skilled in the art based on these specific embodiments without inventive effort are within the scope of protection of this invention.

[0071] To further understand the invention's content, features, and effects, the following specific embodiments are provided, along with accompanying drawings. Figure 1 Detailed explanation is as follows:

[0072] Example 1:

[0073] A method for assessing urban low-altitude fall risk based on a three-dimensional spatial grid includes the following steps:

[0074] S1. The low-altitude airspace of the evaluation airspace is rasterized into three-dimensional voxels of the evaluation airspace to obtain the voxel set of the evaluation airspace.

[0075] Furthermore, the specific implementation method of step S1 includes the following steps:

[0076] S1.1. Select the evaluation airspace The altitude range is from 100 m to 600 m above the ground, covering the typical operating altitudes of mainstream drone logistics, urban air traffic (UAM), and inspection aircraft.

[0077] Set voxel side length ,in Let x be the x-axis side length of the voxel. Let be the y-axis side length of the voxel. Let z be the z-axis side length of the voxel. The side length;

[0078] Further input: research scope boundaries Digital elevation model (DEM), 3D building model and usage attributes, airspace restriction / no-fly zone vectors, and a unified planar projection coordinate system and height datum. To balance accuracy and efficiency, the default... =50m, which can be adjusted to 25m (high-precision urban area) or 100m (wide-area coarse assessment) according to the actual application scenario.

[0079] S1.2. Divide the evaluation spatial domain into sets of voxels at equal intervals. ,in, For voxels, for Any one of them, To determine the number of voxels that divide the study region on the x-axis, for Any one of them, To determine the number of voxels that divide the study region along the y-axis, for Any one of them, The number of voxels that divide the study area along the z-axis;

[0080] Set the voxel center coordinates to ,in Here are the coordinates of the starting reference point on the x-axis for the study area. Let the coordinates of the starting reference point on the y-axis be the coordinates of the study area. The coordinates of the starting reference point of the study area on the z-axis;

[0081] Set row-major one-dimensional encoding , It is a ternary index;

[0082] S1.3. Perform a voxel validity assessment. If a voxel intersects with a building entity, a no-fly zone, or a terrain obstacle, it is marked as invalid; otherwise, it is valid. At the same time, based on management requirements, extend the building horizontally outward by 10-20 m to form a safety buffer zone to obtain valid voxels.

[0083] Furthermore, the horizontal expansion of the building A safety buffer zone (e.g., 10-20 m) can be formed, and stricter buffer values ​​can be used for specific facilities (e.g., schools, hospitals). The voxels are used in subsequent environmental assignment, trajectory crossing and risk calculation.

[0084] S1.4. Spatial indexing and topological association are performed using 6-adjacent-face adjacency based on effective voxels. S2. The building shelter formed by population density grid or mobile signaling inversion population, sensitive facility vector, and 3D building model is quantified and assigned to the voxels of the evaluation spatial domain obtained in step S1, resulting in the voxel set of the evaluation spatial domain;

[0085] Furthermore, an adjacency list is maintained for each voxel to facilitate connectivity and path search; DDA (Digital Differential Analyzer) step parameters for trajectories / rays and voxel meshes are pre-calculated, and a jump table for "voxel index below the entrance face" is stored. If a voxel intersects with a building entity, a no-fly zone, or a terrain obstacle, it is marked as invalid; otherwise, it is valid.

[0086] S2. Quantify the population inversion population, sensitive facility vector, tree canopy shading, and building shelter formed by the three-dimensional building model from the population density grid or mobile signaling, and assign the values ​​to the voxels of the evaluation spatial domain obtained in step S1 to obtain a voxelized environmental parameter database covering the entire domain.

[0087] Objective: To quantify surface population exposure, sensitive targets, and shelter / protection capabilities, and assign values ​​to voxels. , serving as the surface constraint for subsequent layer dependency calculations.

[0088] Input: Population density grid or population inversion from mobile signaling Sensitive facility vector Three-dimensional architectural models form architectural shelters Sentinel-2 L2A multi-temporal imagery is used for remote sensing masking; DEM / DSM is used for surface benchmark consistency.

[0089] Furthermore, the specific implementation method of step S2 includes the following steps:

[0090] S2.1. Perform data alignment and rasterization to obtain the corresponding mobile signaling inversion population. Sensitive facility vector Three-dimensional architectural models form architectural shelters ;

[0091] Furthermore, by unifying coordinates, resolution, and time periods, day / night or multi-time period versions of the population layer can be generated as needed.

[0092] S2.2. Computational Fusion and Exposure, expressed as:

[0093]

[0094]

[0095] in, For integration and Exposure Population inversion for interval-normalized mobile signaling, For interval-normalized sensitive facility vectors, The weights for the population derived from inter-regional normalized mobile signaling. The weights of the interval-normalized sensitive facility vector;

[0096] S2.3. Collect the building shelter coefficient, which is obtained by empirical regression of building volume, height, and materials. ;

[0097] Calculate tree canopy shading coefficient: Calculate Sentinel-2 spectral coefficient NDVI to estimate canopy coverage. The tree canopy shading coefficient is obtained by comparing the mean NDVI with the mean NDVI. :

[0098]

[0099] in, The mean of the Sentinel-2 spectral coefficients. These are empirical parameters;

[0100] Calculate the shading factor The resulting expression is:

[0101] ;

[0102] S2.4. Voxelize the exposure and shading coefficients to obtain the voxelized exposure. and the occlusion coefficient after voxel assignment The expression is:

[0103]

[0104]

[0105] in, This represents the average exposure level.

[0106] S2.5. Divide the spatial domain corresponding to the raster into cubic voxel grids, and integrate all the information obtained in steps S2.1 to S2.5 in the raster to obtain a voxelized environmental parameter database covering the entire domain.

[0107] S3. Construct an aircraft failure rate model;

[0108] Objective: To quantify the flight vehicle's performance at different altitudes. Failure rate per unit time during runtime , used for voxel-level risk synthesis.

[0109] Inputs: Model reliability / Mean Time Between Failures (MTBF); Failure Mode and Effects Set (Power interruption, loss of control, energy depletion, etc.); Mission profile .

[0110] Furthermore, the specific implementation method of step S3 includes the following steps:

[0111] S3.1. Collect the aircraft's model mean time between failures (MTBF) and failure mode set. The failure modes include power interruption, loss of control, and energy depletion. (Task profile) Determine the set of failure modes The failure mode weights are used to query the base failure rate for each failure mode. ;

[0112] S3.2. Constructing the aircraft failure rate model yields:

[0113]

[0114]

[0115] in, For aircraft failure rate, Let m be the weight of the m-th failure mode. Let m be the base failure rate for the m-th failure mode. The task profile weight function, It represents the center height of the k-th height layer.

[0116] S4. Based on the aircraft failure altitude obtained in step S3, conduct an uncertainty simulation test of the aircraft's fall trajectory to obtain the aircraft's landing coverage probability and terminal velocity;

[0117] Furthermore, the simulation input for step S4 is the wind field profile parameters and gust intensity. Based on the relevant duration T, aerodynamic parameter range, initial attitude / velocity range, building / terrain voxels and effective voxel domain, the system is set to terminate upon ground contact and records end-effector velocity samples. and layer influence area The hit count is obtained by convolving the landing kernel of each sample with the ground grid and then normalizing it into a landing coverage probability density distribution. .

[0118] Furthermore, the objective is: targeting "from the layer" In the case of "failure," considering the uncertainties of wind field and attitude / drag, the probability density distribution of landing coverage is obtained. With terminal velocity .

[0119] Input: Wind field profile parameters (power law / logarithmic law) ), gust intensity With relation to duration T; aerodynamic parameter range Initial attitude / velocity range Building / terrain voxels and effective voxel domains.

[0120] The process is as follows:

[0121] S4.1. Sample Generation: ;

[0122] S4.2. Stochastic Differential Dynamics Integral:

[0123]

[0124]

[0125]

[0126] Among them, OU gust dispersion:

[0127]

[0128] The numerical solution adopts the Euler-Maruyama / RK2 numerical method, with a time step Δt∈[0.02,0.1] s;

[0129] S4.3. Termination and Collision Decision: Ground Contact That is, terminate; record the terminal velocity sample. ;

[0130] S4.4. Terminal velocity representative value and influence kernel: A stratified representative terminal velocity based on sample statistics. Or, using an approximate formula:

[0131]

[0132] ;

[0133] Define the radius and area of ​​influence:

[0134]

[0135] ;

[0136] S4.5. Landing Coverage Probability: Convolve the landing kernel of each sample with the ground grid to obtain the hit count. And unified as:

[0137]

[0138] ;

[0139] S4.6. Convergence control: The stopping criterion is that the coefficient of variation of the coverage probability CV ≤ 5% or the L1 distance before and after doubling the sample ≤ ε (e.g., 0.01).

[0140] S5. Combine the voxelized environmental parameter database covering the entire domain obtained in step S2 with the aircraft failure rate obtained in step S3 and the aircraft crash trajectory obtained in step S4, and consider the probability of collision with people to perform voxel risk synthesis to obtain voxel risk.

[0141] Objective: In voxels By integrating the stages of "event occurrence - impact coverage - collision with a person - resulting in injury or death", a risk value is obtained. .

[0142] Input: The output of step two Step 3 Step four .

[0143] Furthermore, the specific implementation method of step S5 includes the following steps:

[0144] S5.1. Calculate the probability of colliding with a person, using the following expression:

[0145]

[0146] Among them, is the probability of collision with people, is the population density, is the layer influence area;

[0147] S5.2. Calculate the casualty probability, and the expression is:

[0148]

[0149]

[0150]

[0151] Among them, is the casualty probability, is the kinetic energy at the end of the layer; are the 0th fatality rate parameter, the 1st fatality rate parameter, and the 2nd fatality rate parameter respectively; is the mass of the drone, is the S-shaped logic function;

[0152] S5.3. Perform voxel risk synthesis to obtain the risk of the voxel , and the expression is:

[0153] .

[0154] S6. Perform risk grading on the risk of the voxel obtained in step S5 to complete a risk assessment of urban low-altitude ground impact based on a three-dimensional space grid.

[0155] Purpose: Organize voxel-level risks into data products available for supervision and planning, and support risk slicing at different altitude levels and route feasibility assessment.

[0156] Input: Voxel risk field { }; grading threshold or target safety level (TLS); aircraft type / time period screening conditions (optional).

[0157] Furthermore, the specific implementation method of step S6 is to set the first threshold as T1, the second threshold as T2, and the third threshold as T3,

[0158] Threshold T1 < T2 < T3, perform risk grading on the risk of the voxel obtained in step S5, and obtain:

[0159]

[0160] Among them, is the risk level of the voxel.

[0161] Furthermore, set up a three-dimensional database and layers: output to the spatio-temporal database; by any Generate two-dimensional slices and (green / yellow / red) suitability maps. The abbreviations of this invention are as follows:

[0162] GIS (Geographic Information System): A technological system used for collecting, managing, and analyzing geospatial data.

[0163] DEM (Digital Elevation Model): A digital model that reflects the distribution of terrain elevation.

[0164] ICAO (International Civil Aviation Organization): A specialized agency of the United Nations responsible for developing international civil aviation standards and regulations.

[0165] EASA (European Union Aviation Safety Agency): The agency responsible for aviation safety oversight and the development of airworthiness standards within the European Union.

[0166] SORA (Specific Operations Risk Assessment): A risk assessment framework for drone and urban air traffic operations adopted in Europe and several other countries.

[0167] OU (Ornstein–Uhlenbeck stochastic process): A stochastic process model used to describe gust disturbances.

[0168] TLS (Target Level of Safety): A metric used in aviation safety to measure acceptable levels of risk.

[0169] Voxel: A regular cubic unit divided in three-dimensional space. This invention uses a default resolution of 50 m × 50 m × 50 m, which can be adjusted as needed.

[0170] 3D building model: contains three-dimensional data of a building's geometry, location, and usage attributes, used to determine whether a fall point provides shelter or causes damage to the building.

[0171] Failure Rate: The probability that an aircraft will fail within a unit of flight time or flight hour.

[0172] Crash trajectory probability: The probability of the landing point distribution on the ground after an aircraft fails, due to the influence of wind field and attitude disturbances.

[0173] Population Exposure: The number or density of people exposed to the risk of falling to the ground per unit area or unit volumetric within a certain time window.

[0174] Shielding Factor: A correction factor introduced to consider the protective effect of structures such as buildings and tree canopies on the risks to people on the ground.

[0175] Collision Probability: The spatial intersection probability resulting from the superposition of the aircraft crash cover core and the ground population distribution.

[0176] Injury / Fatality Probability: The probability of injury or death determined by the terminal impact kinetic energy.

[0177] Three-dimensional Risk Volume: A three-dimensional risk field obtained by superimposing failure rate, trajectory distribution, population exposure and injury probability within a unified spatial grid framework.

[0178] Risk grading: Voxel-level risk values ​​are mapped to red / yellow / green levels according to thresholds or target security levels (TLS) for planning and regulatory visualization.

[0179] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0180] Although this application has been described above with reference to specific embodiments, various modifications can be made and components can be replaced with equivalents without departing from the scope of this application. In particular, as long as there is no structural conflict, the features in the specific embodiments disclosed in this application can be combined with each other in any way. The lack of an exhaustive description of these combinations in this specification is merely for the sake of brevity and resource conservation. Therefore, this application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A method for assessing urban low-altitude fall risk based on a three-dimensional spatial grid, characterized in that, It includes the following steps: S1. Perform three-dimensional rasterization on the low-altitude airspace of the evaluation airspace to obtain voxels of the evaluation airspace, and obtain a voxel set of the evaluation airspace; S2. Quantify the population density grid or mobile signaling-inverted population, sensitive facility vectors, tree canopy occlusion, and building shelters formed by three-dimensional building models, and assign them to the voxels of the evaluation airspace obtained in step S1 to obtain a voxelized environmental parameter database covering the entire region; S3. Construct an aircraft failure rate model; S4. Based on the aircraft failure height in step S3, conduct an uncertainty simulation test on the aircraft fall trajectory to obtain the landing coverage probability density distribution and terminal velocity of the aircraft; S5. Combine the voxelized environmental parameter database obtained in step S2, the aircraft failure rate obtained in step S3, and the aircraft fall trajectory obtained in step S4, consider the probability of collision with people, and perform voxel risk synthesis to obtain the risk of voxels; S6. Perform risk grading on the risks of the voxels obtained in step S5 to complete an urban low-altitude landing risk assessment based on three-dimensional space grids.

2. The urban low-altitude fall risk assessment method based on a three-dimensional spatial grid according to claim 1, characterized in that, The specific implementation method of step S1 includes the following steps: S1.

1. Select the evaluation airspace The altitude range is from 100 m to 600 m above the ground, covering the typical operating altitudes of mainstream drone logistics, urban air traffic (UAM), and inspection aircraft. Set voxel side length ,in Let x be the x-axis side length of the voxel. Let be the y-axis side length of the voxel. Let z be the z-axis side length of the voxel. The side length; S1.

2. Divide the evaluation spatial domain into sets of voxels at equal intervals. ,in, For voxels, for Any one of them, To determine the number of voxels that divide the study region on the x-axis, for Any one of them, To determine the number of voxels that divide the study region along the y-axis, for Any one of them, The number of voxels that divide the study area along the z-axis; Set the voxel center coordinates to ,in Here are the coordinates of the starting reference point on the x-axis for the study area. Let the coordinates of the starting reference point on the y-axis be the coordinates of the study area. The coordinates of the starting reference point of the study area on the z-axis; Set row-major one-dimensional encoding , It is a ternary index; S1.

3. Perform voxel validity judgment. If the voxel intersects with building entities, no-fly / restricted areas or terrain obstacles, it is marked as invalid, otherwise it is valid; at the same time, according to management requirements, expand the building horizontally by 10-20 m to form a safety buffer zone to obtain valid voxels; S1.

4. Use 6-neighbor adjacency to perform spatial indexing and topological association based on valid voxels.

3. The urban low-altitude fall risk assessment method based on a three-dimensional spatial grid according to claim 2, characterized in that, The specific implementation method of step S2 includes the following steps: S2.

1. Perform data alignment and rasterization to obtain the corresponding mobile signaling inversion population. Sensitive facility vector Three-dimensional architectural models form architectural shelters ; S2.

2. Computational Fusion and Exposure, expressed as: ； ； in, For integration and Exposure Population inversion for interval-normalized mobile signaling, For interval-normalized sensitive facility vectors, The weights for the population derived from inter-regional normalized mobile signaling. The weights of the interval-normalized sensitive facility vector; S2.

3. Collect the building shelter coefficient, which is obtained by empirical regression of building volume, height, and materials. ; Calculate tree canopy shading coefficient: Calculate Sentinel-2 spectral coefficient NDVI to estimate canopy coverage. The tree canopy shading coefficient is obtained by comparing the mean NDVI with the mean NDVI. : ； in, The mean of the Sentinel-2 spectral coefficients. These are empirical parameters; Calculate the shading factor The resulting expression is: ； S2.

4. Voxelize the exposure and shading coefficients to obtain the voxelized exposure. and the occlusion coefficient after voxel assignment The expression is: ; ； in, This represents the average exposure level. S2.

5. Divide the airspace corresponding to the grid into cubic voxel grids, and integrate all the information obtained in steps S2.1-S2.5 in the grid to obtain a voxelized environmental parameter database covering the entire region.

4. The urban low-altitude fall risk assessment method based on a three-dimensional spatial grid according to claim 3, characterized in that, The specific implementation method of step S3 includes the following steps: S3.

1. Collect the aircraft's model mean time between failures (MTBF) and failure mode set. The failure modes include power interruption, loss of control, and energy depletion. (Task profile) Determine the set of failure modes The failure mode weights are calculated based on aircraft type airworthiness data and historical operational records to obtain the base failure rate for each failure mode. ; S3.

2. Construct an aircraft failure rate model to obtain: ； ； in, For aircraft failure rate, Let m be the weight of the m-th failure mode. Let m be the base failure rate for the m-th failure mode. The task profile weight function, It represents the center height of the k-th height layer.

5. The urban low-altitude fall risk assessment method based on a three-dimensional spatial grid according to claim 4, characterized in that, Step S4 simulation inputs are wind field profile parameters and gust intensity. Based on the relevant duration T, aerodynamic parameter range, initial attitude / velocity range, building / terrain voxels and effective voxel domain, the system is set to terminate upon ground contact and records end-effector velocity samples. and layer influence area The hit count is obtained by convolving the landing kernel of each sample with the ground grid and then normalizing it into a landing coverage probability density distribution. .

6. The urban low-altitude fall risk assessment method based on a three-dimensional spatial grid according to claim 5, characterized in that, The specific implementation method of step S5 includes the following steps: S5.

1. Calculate the probability of collision with people, and the expression is: ； in, The probability of colliding with a person. For population density, The area affected by the layer; S5.

2. Calculate the casualty probability, and the expression is: ； ； ； in, For the probability of injury or death, As terminal kinetic energy; These are the baseline lethality parameter, energy sensitivity parameter, and shelter effect parameter, respectively. For the quality of drones, It is an S-type logic function; S5.

3. Perform voxel risk synthesis to obtain the voxel risk. The expression is: 。 7. The urban low-altitude fall risk assessment method based on a three-dimensional spatial grid according to claim 6, characterized in that, The specific implementation method of step S6 is to set the first threshold as T1, the second threshold as T2, and the third threshold as T3, where the threshold T1 < T2 < T3, and perform risk grading on the risks of the voxels obtained in step S5 to obtain: ； in, Risk level of voxels.

Images