Urban low-altitude take-off and landing hub dynamic capacity evaluation and air-ground collaborative scheduling method and system

Abstract

The application provides a city low-altitude take-off and landing hub dynamic capacity evaluation and air-ground collaborative scheduling method and system, the method comprises the following steps: constructing a low-altitude take-off and landing hub scene operation network model, and abstracting the network model into a weighted directed graph; obtaining micro-weather data based on a weather sensor, and constructing a local wind field state perception model; establishing a take-off and landing area dynamic availability determination model which fuses safety boundaries and aerodynamic constraints, and updating node states and the network in real time; constructing a path optimization model which considers sliding distance, time, conflict risk and availability, recalculating the operation path and calculating the dynamic capacity; constructing a scheduling system which comprises a weather perception module, an availability determination module, a path optimization module, a capacity evaluation module, an adaptive scheduling module and a visual monitoring module, generating scheduling instructions such as flow limiting according to the matching relationship between the dynamic capacity and the demand, and realizing integrated decision support.

Description

Technical Field

[0001] This invention relates to the field of low-altitude airspace operation management and collaborative scheduling technology, specifically to a method and system for dynamic capacity assessment and air-ground collaborative scheduling of urban low-altitude take-off and landing hubs. Background Technology

[0002] With the deepening development of the low-altitude economy and the acceleration of infrastructure construction, it is foreseeable that urban air traffic and drone logistics will experience explosive growth. In high-density urban areas with hundreds of planned take-off and landing sites, the take-off and landing hub (Vertihub) serves as the core physical support node, and its operational efficiency directly determines the throughput limit of the entire region's low-altitude flight network.

[0003] Existing low-altitude takeoff and landing management platforms typically use ideal static capacity models for calculation, assuming that a certain level of takeoff and landing hub has a fixed number of takeoffs and landings per hour, without fully considering the endogenous impact of the micro-topology of the takeoff and landing field, such as the layout design of peripheral or central taxiways and the parking position ratio, on the actual operating efficiency.

[0004] Furthermore, the meteorological conditions in urban high-rise environments are complex, and transient changes in wind speed and direction have a decisive impact on the takeoff and landing safety of vertical takeoff and landing (VTOL) aircraft. Traditional static evaluation models neglect aerodynamic constraints; for example, aircraft are prone to entering dangerous vortex loop states during takeoff and landing under tailwind conditions, while crosswinds exceeding thresholds can severely affect flight attitude control. In actual operation, changes in the wind field can force the closure of certain takeoff and landing zones or restrict them to one-way operation, forcing aircraft to navigate longer surface taxiways to find available takeoff and landing points, thus significantly increasing the time required for a single dispatch.

[0005] This assessment method, which disconnects macroscopic networks from microscopic topology and separates static assumptions from dynamic weather, makes it impossible for the system to accurately predict changes in the actual capacity of takeoff and landing areas when faced with sudden weather changes. This not only easily leads to congestion and circling waiting of multiple aircraft in the airspace, increasing the risk of airspace collisions and energy consumption, but also results in a lack of accurate and real-time underlying physical data support for macroscopic scheduling.

[0006] Therefore, there is an urgent need in this field for an intelligent capacity assessment method that can deeply integrate dynamic meteorological perception, surface geometric micro-topology, and physical scheduling execution, in order to break through the technical bottlenecks in the operation and management of existing low-altitude infrastructure and ensure the operational efficiency and safety of high-density urban take-off and landing hub networks. Summary of the Invention

[0007] Purpose of the invention: The technical problem to be solved by the present invention is to address the shortcomings of the existing technology by providing a method and system for dynamic capacity assessment and air-ground coordinated scheduling of urban low-altitude take-off and landing hubs. This system enables real-time perception of meteorological constraints of low-altitude take-off and landing fields, dynamic optimization of surface taxiing topology paths, and accurate calculation of actual throughput capacity. It provides a scientific basis for congestion prevention and control of high-density urban low-altitude take-off and landing networks, issuance of adaptive scheduling instructions, and design and planning of low-altitude infrastructure.

[0008] The method includes the following steps:

[0009] Step 1, Construct the surface operation network topology model: The physical space of the low-altitude take-off and landing hub is analyzed into parking position nodes, take-off and landing area nodes and taxiway edges, and the physical attributes and operation weight parameters of the nodes and edges are determined. A weighted directed graph structure reflecting the spatial connectivity is constructed as the surface operation network topology model.

[0010] Step 2, construct a micro-meteorological state perception model: based on the meteorological sensor array deployed in the boundary area of ​​the take-off and landing hub, real-time transient wind direction and wind speed vector data are collected, and smoothed by the edge computing filtering algorithm to generate an effective wind field reference sequence for characterizing local wind field features;

[0011] Step 3, construct the dynamic availability determination model of the take-off and landing area: input the effective wind field reference sequence into the determination model that integrates the aerodynamic constraints of the aircraft and the static building obstruction boundary, calculate the safe entry and exit sector of each take-off and landing area node under the current micro-meteorological conditions in real time, and refresh the dynamic availability status identifier of the corresponding node.

[0012] Step 4: Construct a capacity optimization model that integrates collision avoidance penalties: Based on the updated availability status of take-off and landing area nodes, dynamically adjust the congestion and collision avoidance penalty weights in the field operation network topology model; call the shortest path algorithm to replan the collision avoidance taxiing route from the parking position to the available take-off and landing area, and combine loading / unloading and dwell time to calculate the dynamic theoretical capacity limit of the take-off and landing hub within the preset time window.

[0013] Step 5, Perform capacity matching and adaptive collaborative scheduling: Compare the expected traffic flow in and out demand with the dynamic theoretical capacity limit in real time. When it is determined that the demand is overloaded, automatically generate and issue scheduling instructions based on priority, including physical locking and delayed pushback of blocked aircraft stands and flow restriction and redirection of air routes.

[0014] Step 1 includes:

[0015] The surface operation network topology model is represented as follows: ,in, Represents the set of scene nodes. The set of edges of the sliding channel; the set of scene nodes From the subset of take-off and landing zone nodes subset of parking position nodes constitute, The sliding channel side set From the main taxiway side subset and secondary skid edge subset constitute, The topology selection for the aforementioned network topology model includes either a peripheral layout or a central layout.

[0016] If an external layout is adopted, the main taxiway is laid out along the physical perimeter of the take-off and landing hub and directly connected to the take-off and landing area nodes located at the corner of the station. The parking position nodes are distributed inside the main taxiway and connected to the main taxiway through secondary taxiways.

[0017] If a central layout is adopted, the main taxiway runs through the geometric center of the take-off and landing hub in a cross or radial pattern, and each take-off and landing zone node is located at the end of the main taxiway, while the parking position nodes are symmetrically distributed along both sides of the main taxiway.

[0018] Step 2 includes the following steps:

[0019] Step 2-1: Using a high-frequency meteorological sensor array deployed outside the physical boundary of the low-altitude take-off and landing hub and at the edges of each take-off and landing area, the original wind field state vector, which includes the natural wind field and the high-frequency downwash airflow interference at the moment of aircraft take-off and landing, is collected in real time in discrete time series. ,in For transient wind speed, This refers to the transient wind direction angle;

[0020] Step 2-2: Construct the Kalman filter state smoothing function based on edge computing. The original wind field state vector is used as the observation input, combined with the set system observation noise covariance parameter. Dynamically filtering out high-frequency local instantaneous downwash airflow noise is expressed as: ;in It is the effective wind field reference sequence after smoothing and noise reduction;

[0021] Steps 2-3: Output the smoothed and denoised effective wind field reference sequence. , serving as the time-varying environmental input variable for subsequent aerodynamic constraint boundary determination and scene topology reconstruction.

[0022] Step 3 includes the following steps:

[0023] Step 3-1, Perform tailwind stall boundary determination based on longitudinal aerodynamic projection: Establish a relative airframe coordinate system and map the effective wind field reference sequence to the target aircraft's relative takeoff and landing area nodes. Expected heading angle for approach and departure maneuvers On the vertical axis, calculate the longitudinal along-wind aerodynamic projection component of the current transient wind field on the fuselage's longitudinal axis. :

[0024] ,

[0025] If and only if the longitudinal along-wind aerodynamic projection component Satisfy physical constraints When the course interval is determined to meet the safety control requirements for preventing tailwind stall, the set of compliant course angle intervals is defined as the longitudinal safety set. ;

[0026] Step 3-2: Perform crosswind extremum determination based on lateral quadratic energy projection: extract the lateral aerodynamic projection components of the effective wind field on the fuselage transverse axis perpendicular to the expected heading. :

[0027] ,

[0028] To eliminate the nonlinear nondifferentiability defect of absolute value non-smooth functions in trajectory planning of the underlying control system, a maximum crosswind safe operation threshold based on the aircraft performance envelope is introduced. Construct crosswind quadratic energy boundary constraints:

[0029] ,

[0030] Define the set of compliant heading angle intervals that satisfy the crosswind quadratic energy boundary constraints as the lateral safety set. ;

[0031] Step 3-3, Perform state metric mapping and availability discrete update: Solve for the longitudinal security set. With lateral security sets The intersection of these points generates a set of feasible near-path arcs for the wind field. Set of feasible exit path arcs for wind field ;Will and Each is related to the set of approach path arcs that are statically obstructed by the inherent buildings of the hub. and static exit path arc set Perform intersection mapping to obtain the synthetic feasible near arc. Combined with feasible exit arc ; Calculate the sum of the effective angles of the combined feasible approach arc and the combined feasible departure arc, and compare it with the minimum feasible sector angle threshold. Real-time updates of the state vectors of all nodes in the take-off and landing areas. .

[0032] Step 4 includes the following steps:

[0033] Step 4-1: Construct a scene composite weight matrix that integrates dynamic congestion and entity collision avoidance penalties: Run a network topology model for the scene. Any sliding channel edge in ,in, Indicates the connection node With nodes The gliding path, , Node numbering for parking positions or takeoff and landing areas; introducing standard ground taxiing speeds for UAVs. Spatial length of physical edge Calculate the base gliding time; introduce a node attribute collision avoidance penalty. Dynamic cross-traffic congestion penalty items Construct a composite dynamic weight function :

[0034] ,

[0035] target node When the target node is detected When in a protected state performing loading, unloading, or charging operations, an infinite penalty is imposed. By forcibly blocking the path through the active parking position, and taking action against unprotected nodes. ;

[0036] Targeting dynamic cross-traffic congestion penalties Preset safe passage capacity threshold for the taxiway side It also provides real-time statistics on the estimated traffic flow to be allocated to the taxiway. The safe passage capacity threshold is related to the size of the takeoff and landing area. For example, if the physical side length of a main taxiway is 18 meters, and the minimum queuing distance for a target eVTOL aircraft, including the safety buffer zone, is set to 6 meters, then the safe passage capacity threshold for that taxiway segment is quantified as follows: Flights; when traffic flow When the capacity is below the safe passage threshold, set When traffic flow When the safe passage capacity threshold is greater than or equal to the threshold, the non-linear increase occurs. To guide subsequent drones to detour, the expression is:

[0037] ,

[0038] in, The basic penalty impedance coefficient set for the system, A nonlinear penalty exponent greater than 1 is used to generate a time-varying global weight matrix for the entire field. ;

[0039] Step 4-2, perform bidirectional gliding path topology optimization under time-varying constraints: in the global weight matrix State vectors of all takeoff and landing zone nodes Based on this, the Dijkstra algorithm from the existing shortest path search algorithm is invoked to perform bidirectional optimization for each connected parking position node to decouple entry and exit; the average comprehensive taxiing time of all effective parking position nodes in the field is calculated based on the optimization calculation results. ;

[0040] Step 4-3, construct a dynamic operational capacity measurement model based on air-ground dual bottleneck constraints: set the standard time window for macro-scheduling as... The standard turnaround time for a single parking position is The physical dwell time for a single aircraft occupying the takeoff and landing area for landing or takeoff is Remove isolated machine station nodes that cannot be connected due to the infinite penalty, and count the total number of validly connected parking positions across the entire site. Extract the current state vector Number of nodes in the effective take-off and landing area The parking space limit capacity based on surface turnover efficiency was calculated respectively. Compared to takeoff and landing zone capacity restrictions based on airspace frequency limits :

[0041] ,

[0042] ,

[0043] Pick and The minimum value of the intersection is output as the dynamic operating capacity of the city's low-altitude take-off and landing hub within the current time window. .

[0044] Step 5 includes:

[0045] Extract the set of approach drone sorties and the set of pushback takeoff demand sorties within the time slot, and sum them to generate the total expected demand. Construct a capacity determination threshold margin function. Real-time assessment of the site's load-bearing capacity;

[0046] When the overload threshold is triggered At that time, the low-altitude takeoff and landing management platform calculates the dynamic landing priority weight ranking based on the aircraft's remaining energy status and mission urgency, and generates and issues a set of physical scheduling instructions, including front-end flow restriction and surface blocking, to multi-level linkage agencies in the airspace.

[0047] Between step 4 and step 5, a digital twin 3D visualization monitoring step is further included:

[0048] The aforementioned field operation network topology model, effective wind field reference sequence, state vectors of all take-off and landing zone nodes, and taxiing path topology optimization results are uniformly mapped into three-dimensional graphics pipeline data and input into a three-dimensional graphics engine based on a web-based graphics rendering architecture; a 3D twin availability topology view containing static building occlusion shadow effects, dynamic wind flow field particle effects, and field safety taxiing guiding optical flow is instantiated and rendered in real time.

[0049] The present invention also provides a dynamic capacity assessment and air-ground coordinated scheduling system for urban low-altitude take-off and landing hubs to implement the method, comprising the following modules:

[0050] The scene operation network topology modeling module is used to construct a weighted directed graph topology model consisting of parking position nodes, take-off and landing area nodes, and taxiway edges;

[0051] The micro-meteorological state perception module is used to collect wind direction and speed in real time through a meteorological sensor array, and output a smoothed effective wind field reference sequence based on an edge computing filtering algorithm.

[0052] The take-off and landing area dynamic availability determination module is used to run the aerodynamic constraint determination model and the static building occlusion mapping, and update the availability status of the take-off and landing area nodes in the discrete state space in real time.

[0053] The capacity optimization and evaluation module that integrates collision avoidance penalties is used to dynamically adjust the global weight matrix of the scene, which includes collision avoidance penalties and congestion penalties, to perform bidirectional shortest path optimization, and to calculate the dynamic operating capacity of the take-off and landing hub based on the dual bottlenecks of the ground and airspace.

[0054] The adaptive collaborative scheduling and monitoring module is used to generate and issue front-end current limiting and field blocking physical control commands based on priority sorting when the load is overloaded, and to perform digital twin monitoring in real time through a 3D rendering interface.

[0055] The present invention also provides an electronic device, including a processor and a memory, the memory storing program code that, when executed by the processor, causes the processor to perform the steps of the method.

[0056] The present invention also provides a storage medium storing a computer program or instructions that, when the computer program or instructions are run on a computer, execute the steps of the method described.

[0057] Beneficial Effects: This invention addresses the challenges of surface congestion and takeoff / landing safety caused by sudden micro-meteorological changes in low-altitude operations within high-density urban areas. It enables dynamic control of the operational capacity of takeoff and landing hubs and the scheduling of unmanned aerial vehicles (UAVs). Compared to existing methods that rely solely on ideal static capacity assumptions and ignore the physical spatial layout of takeoff and landing fields, this invention comprehensively characterizes key capabilities such as wind field aerodynamic constraints, micro-geometric topology, node dynamic availability, shortest taxiing optimization, and system collision avoidance response. This improves the comprehensiveness, objectivity, and accuracy of capacity assessment results, reducing the risk of congestion and misjudgment caused by neglecting micro-meteorological changes and local physical lock-in due to single static indicators. Furthermore, this invention comprehensively considers the impact of short-term wind direction and speed changes, different takeoff and landing hub layouts, and the macro-operational status of low-altitude takeoff and landing platforms on surface efficiency, enhancing the applicability and generalization ability of the assessment and scheduling methods to complex urban high-rise obstruction and extreme weather scenarios. By quantitatively comparing and visualizing the comprehensive performance of different topology schemes and scheduling strategies under dynamic wind field scenarios, this study addresses issues such as unreasonable allocation of low-altitude hub resources, delayed response to sudden weather changes, and fragmented air-ground scheduling. This improves the systematicness, operability, and decision support capabilities of urban low-altitude take-off and landing hub network operation management and capacity assessment. Attached Figure Description

[0058] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0059] Figure 1 This is a flowchart of the method of the present invention.

[0060] Figure 2 This is a schematic diagram of the topology mapping of the take-off and landing hub.

[0061] Figure 3 This is a diagram showing the dynamic wind field constraints and state evolution.

[0062] Figure labeling: 301 - Take-off and landing zone node; 302 - Effective wind direction vector; 303 - Comprehensive feasible approach arc; 304 - Comprehensive feasible departure arc; 305 - Static building obstruction zone; 306 - Minimum safe sector threshold. Detailed Implementation

[0063] like Figure 1 As shown, this invention provides a method for dynamic capacity assessment and air-ground coordinated scheduling of urban low-altitude take-off and landing hubs, including the following steps:

[0064] Step 1: Construct the surface operation network topology model. The physical space of the low-altitude takeoff and landing hub is analyzed into parking position nodes, takeoff and landing area nodes, and taxiway edges. The physical attributes and operational weight parameters of these nodes and edges are determined, and a weighted directed graph structure reflecting spatial connectivity is constructed. Specifically, this includes:

[0065] Step 1-1 involves analyzing the geometric characteristics of the takeoff and landing area, final approach and takeoff area, and safety zone of the low-altitude takeoff and landing hub, clarifying the spatial boundaries and functional positioning of each core area. The specific implementation method is as follows:

[0066] The spatial scope of the low-altitude takeoff and landing hub is delineated, and the hierarchical relationship between various load-bearing surfaces and safety buffer zones in the physical landscape is clarified; physical parameters are quantified based on the aircraft operational envelope. The set of takeoff and landing zone nodes is defined as follows. For any take-off and landing zone node Its minimum physical diameter is set as the maximum rotor wingspan of the aircraft. The minimum diameter of the Final Approach and Takeoff Area (FATO) is defined as not less than 1.5 times the maximum physical size of the aircraft. To support independent operation of multiple nodes, the physical isolation distance between any two adjacent FATO edges must be greater than or equal to a preset safety threshold.

[0067] Steps 1-2 involve parametrically describing elements such as takeoff and landing area nodes, parking position nodes, and taxiway connectivity mechanisms for different takeoff and landing hub facility layout schemes. Figure 2 As shown, a surface operation network topology model reflecting the differences in microscopic surface space and traffic capacity is constructed. The specific implementation method is as follows:

[0068] The microscopic scene of the take-off and landing hub is abstracted as a mathematical set consisting of multiple key nodes and edges, and a weighted directed graph topological model is constructed, represented as:

[0069] ,

[0070] in, Represents the set of scene nodes. This represents the set of edges of the sliding passage. The set of scene nodes... From the subset of take-off and landing zone nodes subset of parking position nodes Composition, that is The sliding channel side set From the main taxiway side subset and secondary skid edge subset Composition, that is Define the connectivity characteristics of the edge sets under different facility layout schemes:

[0071] For the peripheral layout scheme, the characteristic is the main taxiway side set Deployed along the physical periphery of the take-off and landing hub and directly connected to the take-off and landing area nodes located at the corners. Gathering at the side of the secondary taxiway Located inside the main taxiway and connected to the parking position node The central area of ​​the terminal is used as a centralized loading and unloading space; for the central layout scheme, the main taxiway is used as a side assembly point. It runs through the geometric center of the take-off and landing hub in a crisscross pattern, with nodes in each take-off and landing area. Located at the end of the main taxiway, parking position node It is distributed along both sides of the main taxiway.

[0072] Step 2: Construct a micro-meteorological state perception model. Based on a meteorological sensor array deployed in the boundary area of ​​the take-off and landing hub, transient wind direction and wind speed vector data are collected in real time and smoothed using an edge computing filtering algorithm to generate an effective wind field reference sequence for characterizing local wind field features; the specific implementation method is as follows:

[0073] Define the discrete time series as High-frequency meteorological sensor arrays are deployed outside the physical boundary of the take-off and landing hub and at the edges of each take-off and landing area to collect data in real time. The original physical wind field state vector containing the natural wind field and the noise from the aircraft rotor downwash at any given moment:

[0074] ,

[0075] in, For transient wind speed, This refers to the transient wind direction angle;

[0076] Kalman Filter State Smoothing Function Based on Edge Computation The original physical wind field state vector is used as the observation input, combined with a preset system observation noise covariance parameter. The dynamic filtering and smoothing process for high-frequency local downwash interference induced during aircraft takeoff and landing is represented as follows:

[0077] ,

[0078] Output the effective wind field reference sequence after smoothing and noise reduction. This will be used as the environmental input variable for the subsequent dynamic availability determination model of the take-off and landing area.

[0079] Step 3: Construct a dynamic availability determination model for the takeoff and landing area. For example... Figure 3As shown, the effective wind field reference sequence is input into the judgment model that integrates aircraft aerodynamic constraints and static building obstruction boundaries. The safe approach and departure sectors of each takeoff and landing zone node under the current micro-meteorological conditions are calculated in real time, and the dynamic availability status of the corresponding nodes is updated. Specifically, this includes:

[0080] Step 3-1: Perform downwind stall boundary determination based on longitudinal aerodynamic projection. The specific implementation method is as follows:

[0081] Establish a relative body coordinate system and map the effective wind field reference sequence to the target aircraft's relative takeoff and landing area nodes. Expected heading angle for approach and departure maneuvers On the vertical axis;

[0082] Solve for the longitudinal downwind aerodynamic projection component of the current transient wind field on the fuselage longitudinal axis. The expression is:

[0083] ,

[0084] If and only if the longitudinal aerodynamic projection component satisfies the physical constraint condition When the course interval is determined to meet the safety control requirements for preventing tailwind stall, the set of compliant course angle intervals is defined as the longitudinal safety set. .

[0085] Step 3-2: Perform crosswind extremum determination based on lateral quadratic energy projection. The specific implementation method is as follows:

[0086] Extract the lateral aerodynamic projection component of the effective wind field on the fuselage transverse axis perpendicular to the expected heading. The expression is:

[0087] ,

[0088] To eliminate the nonlinear nondifferentiability defect of absolute value non-smooth functions in trajectory planning of the underlying control system, a maximum crosswind safe operation threshold based on the aircraft performance envelope is introduced. Construct crosswind quadratic energy boundary constraints:

[0089] ,

[0090] Define the set of compliant heading angle intervals that satisfy the extremum constraint of this quadratic form as the lateral safety set. .

[0091] Step 3-3: Perform state metric mapping and availability discrete update. The specific implementation is as follows:

[0092] Solve for the vertical security set With lateral security sets The intersection of these points generates a set of feasible near-path arcs for the wind field. Set of feasible exit path arcs for wind field ;

[0093] Will and Each is related to the set of approach path arcs that are statically obstructed by the inherent buildings of the hub. and static exit path arc set Perform intersection mapping to obtain the synthetic feasible near arc. Combined with feasible exit arc ;

[0094] Calculate the sum of the effective angles of the combined feasible approach arc and the combined feasible departure arc. By comparing the minimum operable sector angle threshold The availability of take-off and landing zone nodes is mapped to a discrete space, and the specific state update logic is as follows:

[0095] like and If so, the node is determined to support bidirectional takeoff and landing, and the state variable is updated to... ;

[0096] like and If the node encounters resistance while attempting to leave the field, its state variable is updated to... ;

[0097] like and If the node encounters resistance during approach, its state variable is updated to... ;

[0098] like and If so, the node is determined to be deadlocked by the wind field, and the state variable is updated to... .

[0099] Real-time updates of the state vectors of all nodes in the takeoff and landing areas. This is then used as a constraint control input for subsequent capacity optimization.

[0100] Step 4: Construct a capacity optimization model that integrates collision avoidance penalties. Based on the updated availability status of takeoff and landing area nodes, dynamically adjust the congestion and collision avoidance penalty weights in the surface operation network topology model; call the shortest path algorithm to replan the collision avoidance taxiing route from the parking position to the available takeoff and landing area, and combine loading / unloading and dwell time to calculate the dynamic operating capacity of the takeoff and landing hub within a preset time window; specifically including:

[0101] Step 4-1: Construct a scene composite weight matrix that integrates dynamic congestion and entity collision avoidance penalties. The specific implementation method is as follows:

[0102] For any sliding channel edge in the network topology model of the scene described above. Introducing standard ground taxiing speed for drones Spatial length of physical edge Calculate the base glide time;

[0103] Introducing node attribute collision prevention penalty Dynamic cross-traffic congestion penalty items Construct a composite dynamic weight function :

[0104] ,

[0105] When the system detects the target node When in a protected state performing loading, unloading, or charging operations, an infinite penalty is imposed. By forcibly blocking the path through the active parking position, and taking action against unprotected nodes. ;

[0106] Targeting dynamic cross-traffic congestion penalties Preset safe passage capacity threshold for the taxiway side It also provides real-time statistics on the estimated traffic flow allocated to that side. When the traffic flow is lower than the safe traffic capacity threshold, set... When the traffic flow approaches or exceeds the threshold, the nonlinear increase occurs. To guide subsequent drones to detour, its expression is:

[0107] ,

[0108] in, The basic penalty impedance coefficient set for the system, A nonlinear penalty exponent greater than 1 is used to generate a time-varying global weight matrix for the entire field. .

[0109] Step 4-2: Perform bidirectional glide path topology optimization under time-varying constraints. The specific implementation method is as follows:

[0110] In the global weight matrix State vectors of all takeoff and landing zone nodes Based on this, the Dijkstra algorithm from the existing shortest path search algorithm is called to perform bidirectional optimization for each connected stop node to decouple entry and exit;

[0111] Calculate the shortest time to drive into the target parking position from each available takeoff and landing area, and the shortest time to drive out of the parking position to each available takeoff and landing area.

[0112] After removing isolated machine station nodes that cannot be connected due to the infinite penalty, the total number of validly connected parking positions in the entire venue is counted. ;

[0113] Calculate the average combined taxiing time of all effective parking positions across the entire field. .

[0114] Step 4-3: Construct a dynamic operational capacity calculation model based on the dual bottleneck constraints of air and ground. The specific implementation method is as follows:

[0115] Set the standard time window for macro-scheduling as The standard turnaround time for a single parking position is The physical dwell time for a single aircraft occupying the takeoff and landing area for landing or takeoff is Calculate the parking space limit capacity based on surface turnover efficiency. :

[0116] ,

[0117] Extract the current state vector Central Africa Number of valid take-off and landing zone nodes in (closed) state Calculate the restricted capacity of the takeoff and landing area based on airspace frequency constraints. :

[0118] ,

[0119] Output the minimum value of the intersection of the two values ​​to determine the dynamic operating capacity of the city's low-altitude take-off and landing hub within the current time window. .

[0120] Step 5, perform capacity matching and adaptive collaborative scheduling. The expected traffic flow inflow and outflow demand is compared in real time with the dynamic operating capacity. When overload is detected, priority-based scheduling instructions are automatically generated and issued, including physical locking and delayed pushback of obstructed aircraft positions, and flow restriction and redirection of aircraft within the airspace; specifically including:

[0121] Step 5-1: Establish a multi-source fusion short-term demand forecasting and load margin calculation mechanism. The specific implementation method is as follows:

[0122] Within the set evaluation time window Within the time slot, the set of approach drone sorties and the set of pushback takeoff demand sorties are extracted and summed to generate the total expected demand. ;

[0123] Constructing the capacity decision margin function It is used to assess the surface load-bearing status in real time, and its expression is:

[0124] ,

[0125] When the margin is satisfied If the system determines that there is redundancy in the current takeoff and landing hub's throughput capacity, it will maintain the established flight plan unchanged.

[0126] When the overload threshold is triggered When it is determined that the current hub is physically overloaded due to wind field deterioration or surface queuing, the system automatically activates the air-ground collaborative adaptive scheduling program.

[0127] Step 5-2: Generate and issue priority-based air-ground cooperative adaptive scheduling instructions. The specific implementation method is as follows:

[0128] When the overload threshold is triggered, the load exceeds the actual safe bearing capacity. The low-altitude takeoff and landing management platform calculates dynamic landing priority weights and sorts each aircraft based on its remaining energy status and mission urgency for the sorting of flight surplus traffic flow.

[0129] Based on the sorting results, a set of physical scheduling instructions, including front-end flow limiting and surface blocking, is generated and issued to the multi-level linkage mechanism in the airspace:

[0130] At the takeoff and landing airspace control layer: through the air-to-ground wireless data link communication network, issue a diversion redirection instruction to the secondary approach UAVs that have not yet entered the terminal control area, and change their target waypoint to the surrounding backup takeoff and landing hubs with redundant capacity; or issue a deceleration and hovering waiting instruction to the secondary priority flights ahead, and guide them to the designated air holding layer.

[0131] At the ground control layer: within the take-off and landing hub, a specific parking position node is automatically locked due to sudden weather changes or the connection edge being under high congestion penalty weight, and a delayed pushback command is issued to the UAVs waiting to leave in that parking position.

[0132] Step 5-3: Perform digital twin 3D visualization monitoring. The specific implementation method is as follows:

[0133] The aforementioned surface operation network topology model, effective wind field reference sequence, take-off and landing zone node state vector, and taxiing path topology optimization results are uniformly mapped into three-dimensional graphical pipeline data.

[0134] The pipeline data is input into a 3D graphics engine based on a web-based graphics rendering architecture, which instantiates and renders in real time a 3D twin availability topology view that includes static building occlusion shadow effects, dynamic airflow particle effects, and scene safety sliding guiding light flow.

[0135] Based on the above-described method for dynamic capacity assessment and air-ground coordinated scheduling of urban low-altitude take-off and landing hubs, this invention provides a system for dynamic capacity assessment and air-ground coordinated scheduling of urban low-altitude take-off and landing hubs. This system is used to execute the method logic in the above steps, and the system includes:

[0136] The surface operation network topology modeling module is used to analyze the physical space of low-altitude take-off and landing hubs. It abstracts parking stand nodes, take-off and landing area nodes, and taxiway edges into a weighted directed graph structure, and constructs a surface operation network topology model that reflects spatial connectivity based on different facility layout schemes such as peripheral or central layouts.

[0137] The micro-meteorological state perception module is used to collect transient wind direction and wind speed vector data in real time through a meteorological sensor array deployed in the boundary area of ​​the take-off and landing hub. It calls the Kalman filter state smoothing function based on edge computing to filter out the interference of high-frequency downwash airflow during take-off and landing, and outputs the smoothed effective wind field reference sequence.

[0138] The take-off and landing area dynamic availability determination module is used to receive the effective wind field reference sequence, establish a relative body coordinate system to perform downwind stall boundary determination based on longitudinal aerodynamic projection, extract lateral projection components to perform lateral quadratic energy boundary constraint determination, and combine static building obstruction arc segment calculation to comprehensively measure the drivable departure arc, and update the availability status identifier of the take-off and landing area nodes in discrete space in real time.

[0139] The capacity optimization and evaluation module integrating collision avoidance penalties is used to dynamically adjust the congestion and entity collision avoidance penalty weights in the scene operation network topology model based on the updated availability status identifier, generating a scene composite weight matrix; it calls the Dijkstra algorithm to perform bidirectional shortest path optimization for decoupling entry and exit, and calculates the dynamic operation capacity limit within a preset time window based on the dual bottleneck constraints of surface turnover and airspace frequency.

[0140] The adaptive collaborative scheduling and monitoring module is used to construct a capacity judgment margin function and compare the total expected demand with the dynamic operating capacity limit in real time. When the judgment margin triggers the overload threshold, it automatically generates a scheduling instruction set based on landing priority weight, issues front-end flow limiting and alternate landing redirection instructions to the airspace control layer, issues physical locking instructions for obstructed aircraft positions to the surface control layer, and performs digital twin visualization rendering through a 3D graphics engine.

[0141] To verify the effectiveness of the assessment and scheduling method described in this invention, in actual urban low-altitude operating environments, the evolution of meteorological conditions often manifests as a continuous and gradual deflection of wind direction. This gradual wind direction deflection causes a continuous, step-like evolution of the aerodynamic availability of nodes in each take-off and landing area. This embodiment uses a low-altitude take-off and landing hub with a periphery layout in a high-density urban area as the simulation object. A gradual meteorological scenario is set where the wind speed remains stable while the wind direction continuously deflects by 105° within 90 minutes. This verifies the real-time tracking capability of the method described in this invention for capacity step-like decay under continuous meteorological evolution conditions, its ability to accurately capture scheduling critical points, and the influence of node spatial orientation differences on the timing of capacity decay under a periphery layout. The parameters are shown in Table 1.

[0142] Table 1

[0143]

[0144] The meteorological parameters were set as follows: wind speed was constant at 10 kt, and wind direction continuously deflected clockwise from an initial 225° at a rate of 17.5° / 15 min, until the end of the 6th time window when the wind direction had cumulatively deflected to 330°. According to this invention, within each time window, downwind stall boundary determination and crosswind extreme value determination were performed on each take-off and landing zone node, and mapped to the discrete state space through a metric function. Tables 2 and 3 show the complete state evolution process of each node within the 6 time windows.

[0145] Table 2

[0146]

[0147] Table 2

[0148] Note: Crosswind constraint determination: All crosswind components in Table 2 above satisfy the lateral constraints. The main reason for node closure is the downwind stall constraint.

[0149] Table 3

[0150]

[0151] Based on the node state vectors updated within each time window, the system uses the Dijkstra algorithm to replan the shortest anti-collision paths from each parking position to the available takeoff and landing area, and counts the number of effectively connected parking positions across the entire field. and average combined gliding time The details are shown in Tables 4 and 5.

[0152] Table 4

[0153]

[0154] Table 5

[0155]

[0156] To simulate stable demand during peak periods, this embodiment sets the expected arrival and departure demand for each time window to remain constant at 32 flights, and the capacity margin function... The calculation and scheduling results are shown in Table 6 below.

[0157] Table 6

[0158]

[0159] In this embodiment, the traditional static ideal capacity of the peripheral layout hub is set at 36 flights / 15 minutes, with a stable expected arrival and departure demand of 32 flights per time slot. During the 0-90 minute simulation period, the system's meteorological sensor array captures a continuous, gradual deflection of 105° in the external wind field, despite a constant wind speed. Because the traditional static assessment model cannot detect the micro-meteorological wind direction evolution, it still outputs an available capacity of 36 flights throughout the entire process. If flights are continuously released according to the traditional plan, [the following will occur]. , Multiple periods of time accumulate to generate a large number of over-releases, posing a serious risk of traffic congestion and mid-air collisions. The dynamic availability determination model for takeoff and landing areas in step 3 of this invention can track wind direction deflection in real time and accurately identify some takeoff and landing area nodes that are physically locked due to exceeding the tailwind stall constraint, thus changing their state. Combined with the shortest taxiing collision avoidance optimization in step 4, the dynamic operational capacity limit of the current time slot was recalculated, resulting in a sharp, step-like drop, falling to a minimum of 20 sorties. Due to... When demand far exceeds capacity limits during certain periods, the system determines... This triggers adaptive scheduling in step 5: redundancy requirements for remaining flights are redirected via air-to-ground data link, held at the standard air deck, or locked to ground positions. The restrictions will be automatically lifted when the wind direction crosses the safety threshold and the capacity recovers.

[0160] Conclusion: Simulation results of the embodiments demonstrate that the present invention can track the multi-step decay of microscopic field capacity caused by progressive wind direction deflection in real time, and capture the critical points for adaptive scheduling triggering and recovery. It can proactively block overloaded traffic flow during periods of weather deterioration, eliminate field deadlock and flight safety hazards caused by inaccurate static assessments, and avoid over-scheduling, thereby achieving precise air-ground coordinated management under continuous weather evolution.

[0161] This invention provides a method and system for dynamic capacity assessment and air-ground coordinated scheduling of urban low-altitude take-off and landing hubs. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment of the invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A method for dynamic capacity assessment and air-ground coordinated scheduling of urban low-altitude take-off and landing hubs, characterized in that, Includes the following steps: Step 1, Construct the surface operation network topology model: The physical space of the low-altitude take-off and landing hub is analyzed into parking position nodes, take-off and landing area nodes and taxiway edges, and the physical attributes and operation weight parameters of the nodes and edges are determined. A weighted directed graph structure reflecting the spatial connectivity is constructed as the surface operation network topology model. Step 2, construct a micro-meteorological state perception model: based on the meteorological sensor array deployed in the boundary area of ​​the take-off and landing hub, real-time transient wind direction and wind speed vector data are collected, and smoothed by the edge computing filtering algorithm to generate an effective wind field reference sequence for characterizing local wind field features; Step 3, construct the dynamic availability determination model of the take-off and landing area: input the effective wind field reference sequence into the determination model that integrates the aerodynamic constraints of the aircraft and the static building obstruction boundary, calculate the safe entry and exit sector of each take-off and landing area node under the current micro-meteorological conditions in real time, and refresh the dynamic availability status identifier of the corresponding node. Step 4: Construct a capacity optimization model that integrates collision avoidance penalties: Based on the updated availability status of take-off and landing area nodes, dynamically adjust the congestion and collision avoidance penalty weights in the field operation network topology model; call the shortest path algorithm to replan the collision avoidance taxiing route from the parking position to the available take-off and landing area, and combine loading / unloading and dwell time to calculate the dynamic theoretical capacity limit of the take-off and landing hub within the preset time window. Step 5, Perform capacity matching and adaptive collaborative scheduling: Compare the expected traffic flow in and out demand with the dynamic theoretical capacity limit in real time. When it is determined that the demand is overloaded, automatically generate and issue scheduling instructions based on priority, including physical locking and delayed pushback of blocked aircraft stands and flow restriction and redirection of air routes.

2. The method according to claim 1, characterized in that, Step 1 includes: The surface operation network topology model is represented as follows: ,in, Represents the set of scene nodes. The set of edges of the sliding channel; the set of scene nodes From the subset of take-off and landing zone nodes subset of parking position nodes constitute, The sliding channel side set From the main taxiway side subset and secondary skid edge subset constitute, The topology selection for the aforementioned network topology model includes either a peripheral layout or a central layout. If an external layout is adopted, the main taxiway is laid out along the physical perimeter of the take-off and landing hub and directly connected to the take-off and landing area nodes located at the corner of the station. The parking position nodes are distributed inside the main taxiway and connected to the main taxiway through secondary taxiways. If a central layout is adopted, the main taxiway runs through the geometric center of the take-off and landing hub in a cross or radial pattern, and each take-off and landing zone node is located at the end of the main taxiway, while the parking position nodes are symmetrically distributed along both sides of the main taxiway.

3. The method according to claim 2, characterized in that, Step 2 includes the following steps: Step 2-1: Using a high-frequency meteorological sensor array deployed outside the physical boundary of the low-altitude take-off and landing hub and at the edges of each take-off and landing area, the original wind field state vector, which includes the natural wind field and the high-frequency downwash airflow interference at the moment of aircraft take-off and landing, is collected in real time in discrete time series. ,in For transient wind speed, This refers to the transient wind direction angle; Step 2-2: Construct the Kalman filter state smoothing function based on edge computing. The original wind field state vector is used as the observation input, combined with the set system observation noise covariance parameter. Dynamically filtering out high-frequency local instantaneous downwash airflow noise is expressed as: ;in It is the effective wind field reference sequence after smoothing and noise reduction; Steps 2-3: Output the smoothed and denoised effective wind field reference sequence. , serving as the time-varying environmental input variable for subsequent aerodynamic constraint boundary determination and scene topology reconstruction.

4. The method according to claim 3, characterized in that, Step 3 includes the following steps: Step 3-1, Perform tailwind stall boundary determination based on longitudinal aerodynamic projection: Establish a relative airframe coordinate system and map the effective wind field reference sequence to the target aircraft's relative takeoff and landing area nodes. Expected heading angle for approach and departure maneuvers On the vertical axis, calculate the longitudinal along-wind aerodynamic projection component of the current transient wind field on the fuselage's longitudinal axis. : ， If and only if the longitudinal along-wind aerodynamic projection component Satisfy physical constraints When the course interval is determined to meet the safety control requirements for preventing tailwind stall, the set of compliant course angle intervals is defined as the longitudinal safety set. ; Step 3-2: Perform crosswind extremum determination based on lateral quadratic energy projection: extract the lateral aerodynamic projection components of the effective wind field on the fuselage transverse axis perpendicular to the expected heading. : ， Introducing a maximum crosswind safe operating threshold based on the aircraft performance envelope. Construct crosswind quadratic energy boundary constraints: ， Define the set of compliant heading angle intervals that satisfy the crosswind quadratic energy boundary constraints as the lateral safety set. ; Step 3-3, Perform state metric mapping and availability discrete update: Solve for the longitudinal security set. With lateral security sets The intersection of these points generates a set of feasible near-path arcs for the wind field. Set of feasible exit path arcs for wind field ;Will and Each is related to the set of approach path arcs that are statically obstructed by the inherent buildings of the hub. and static departure path arc set Perform intersection mapping to obtain the synthetic feasible near arc. Combined with feasible exit arc ; Calculate the sum of the effective angles of the combined feasible approach arc and the combined feasible departure arc, and compare it with the minimum feasible sector angle threshold. Real-time updates of the state vectors of all nodes in the take-off and landing areas. .

5. The method according to claim 4, characterized in that, Step 4 includes the following steps: Step 4-1: Construct a scene composite weight matrix that integrates dynamic congestion and entity collision avoidance penalties: Run a network topology model for the scene. Any sliding channel edge in ,in, Indicates the connection node With nodes The gliding path, , Node numbering for parking positions or takeoff and landing areas; introducing standard ground taxiing speeds for UAVs. Spatial length of physical edge Calculate the base gliding time; introduce a node attribute collision avoidance penalty. Dynamic cross-traffic congestion penalty items Construct a composite dynamic weight function : ， target node When the target node is detected When in a protected state performing loading, unloading, or charging operations, an infinite penalty is imposed. By forcibly blocking the path through the active parking position, and taking action against unprotected nodes. ; Targeting dynamic cross-traffic congestion penalties Preset safe passage capacity threshold for the taxiway side It also provides real-time statistics on the estimated traffic flow to be allocated to the taxiway. When traffic flow When the capacity is below the safe passage threshold, set When traffic flow When the safe passage capacity threshold is greater than or equal to the threshold, the non-linear increase occurs. To guide subsequent drones to detour, the expression is: ， in, The basic penalty impedance coefficient set for the system, A nonlinear penalty exponent greater than 1 is used to generate a time-varying global weight matrix for the entire field. ; Step 4-2, perform bidirectional gliding path topology optimization under time-varying constraints: in the global weight matrix State vectors of all takeoff and landing zone nodes Based on this, the Dijkstra algorithm from the existing shortest path search algorithm is invoked to perform bidirectional optimization for each connected parking position node to decouple entry and exit; the average comprehensive taxiing time of all effective parking position nodes in the field is calculated based on the optimization calculation results. ; Step 4-3, construct a dynamic operational capacity measurement model based on air-ground dual bottleneck constraints: set the standard time window for macro-scheduling as... The standard turnaround time for a single parking position is The physical dwell time for a single aircraft occupying the takeoff and landing area for landing or takeoff is Remove isolated machine station nodes that cannot be connected due to the infinite penalty, and count the total number of validly connected parking positions across the entire site. Extract the current state vector Number of nodes in the effective take-off and landing area The parking space limit capacity based on surface turnover efficiency was calculated respectively. Compared to takeoff and landing zone capacity restrictions based on airspace frequency limits : ， ， Pick and The minimum value of the intersection is output as the dynamic operating capacity of the city's low-altitude take-off and landing hub within the current time window. .

6. The method according to claim 5, characterized in that, Step 5 includes: Extract the set of approach drone sorties and the set of pushback takeoff demand sorties within the time slot, and sum them to generate the total expected demand. Construct a capacity determination threshold margin function. Real-time assessment of the site's load-bearing capacity; When the overload threshold is triggered At that time, the low-altitude takeoff and landing management platform calculates the dynamic landing priority weight ranking based on the aircraft's remaining energy status and mission urgency, and generates and issues a set of physical scheduling instructions, including front-end flow control and surface blocking, to multi-level linkage agencies in the airspace.

7. The method according to claim 6, characterized in that, Between step 4 and step 5, a digital twin 3D visualization monitoring step is further included: The aforementioned field operation network topology model, effective wind field reference sequence, state vectors of all take-off and landing zone nodes, and taxiing path topology optimization results are uniformly mapped into three-dimensional graphics pipeline data and input into a three-dimensional graphics engine based on a web-based graphics rendering architecture. Real-time instantiation and rendering of a 3D twin usability topology view that includes static building occlusion shadow effects, dynamic airflow particle effects, and scene safety glide-guided light flow.

8. A system for dynamic capacity assessment and air-ground coordinated scheduling of urban low-altitude take-off and landing hubs to implement the method described in any one of claims 1 to 7, characterized in that, Includes the following modules: The scene operation network topology modeling module is used to construct a weighted directed graph topology model consisting of parking position nodes, take-off and landing area nodes, and taxiway edges; The micro-meteorological state perception module is used to collect wind direction and speed in real time through a meteorological sensor array, and output a smoothed effective wind field reference sequence based on an edge computing filtering algorithm. The take-off and landing area dynamic availability determination module is used to run the aerodynamic constraint determination model and the static building occlusion mapping, and update the availability status of the take-off and landing area nodes in the discrete state space in real time. The capacity optimization and evaluation module that integrates collision avoidance penalties is used to dynamically adjust the global weight matrix of the scene, which includes collision avoidance penalties and congestion penalties, to perform bidirectional shortest path optimization, and to calculate the dynamic operating capacity of the take-off and landing hub based on the dual bottlenecks of the ground and airspace. The adaptive collaborative scheduling and monitoring module is used to generate and issue front-end current limiting and field blocking physical control commands based on priority sorting when the load is overloaded, and to perform digital twin monitoring in real time through a 3D rendering interface.

9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing program code that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, It stores a computer program or instructions that, when run on a computer, perform the steps of the method as described in any one of claims 1 to 7.

Images