A dynamic allocation management method for low-altitude airspace

Abstract

The application discloses a kind of dynamic allocation management methods for low altitude airspace, to solve the problem of low resource utilization rate of static management mode, lack of dynamic adaptability.The method is first by three-dimensional space-time gridding, airspace is discretized into standardized management unit, constructs parameterization and semantic dynamic airspace unit;Second, establish airspace dynamic allocation mechanism based on multi-objective optimization algorithm, consider airspace utilization rate, conflict risk and other goals;At the same time, hierarchical management strategy is formulated, risk level is dynamically divided by fuzzy comprehensive evaluation, and differential management and control are implemented;Finally, meta-heuristic algorithm is used for solving, realize near real-time, intelligent allocation of airspace resources, form perception-decision-execution-evaluation-optimization adaptive cycle.The application significantly improves the resource utilization rate, safety and system resilience of low altitude airspace, and provides key technical support for the large-scale development of urban air traffic.

Description

Technical Field

[0001] This invention relates to a dynamic allocation and management method for low-altitude airspace, belonging to the field of low-altitude airspace management technology. Background Technology

[0002] The booming development of the low-altitude economy poses a severe challenge to the existing low-altitude airspace management model. Currently, the management approach mainly follows the static airspace delineation concept used for high-altitude civil aviation, that is, pre-delineating fixed airspace structures for long-term use. When faced with a massive, heterogeneous, and highly dynamic number of low-altitude aircraft, this model reveals inherent defects such as low airspace resource utilization efficiency, lack of dynamic adaptability of the system, and coarse management granularity that makes it difficult to achieve refined management.

[0003] First, the rigid isolation between statically defined airspace units prevents flexible adjustments based on real-time mission requirements. Large amounts of airspace resources remain idle when there are no flight missions, while unforeseen demands cannot be met, leading to airspace isolation and hindering the support of high-density low-altitude operations. Second, the low-altitude operating environment is complex and variable, with frequent occurrences of sudden weather events and temporary no-fly zones. The static management model is slow to respond and cannot achieve near-real-time reconstruction of the airspace structure, making it difficult to effectively mitigate operational risks and support emergency response missions, thus becoming a key bottleneck for improving efficiency and safety. Furthermore, the existing model struggles to effectively differentiate and address the mixed operation of aircraft with different performance and mission types. The lack of differentiated control measures for layered and graded airspace management easily leads to a one-size-fits-all approach, restricting the balance between operational dynamism and safety.

[0004] Although the industry has begun to explore dynamic airspace management, existing research mostly remains at the level of simple geometric division, failing to deeply couple with the inherent operational characteristics of airspace, and lacking a complete technical system that integrates dynamic allocation and hierarchical management. Therefore, there is an urgent need to invent a new method for low-altitude dynamic airspace management to achieve real-time intelligent allocation and flexible management of airspace resources, fundamentally improve utilization efficiency, enhance system resilience, and provide technical support for the safe and large-scale development of the low-altitude economy. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a dynamic allocation and management method for low-altitude airspace, which aims to solve the problems of low resource utilization and lack of dynamic adaptability in the static management mode of low-altitude airspace.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: A dynamic allocation and management method for low-altitude airspace includes the following steps: Step S1: Determine the target low-altitude airspace range and time period to be allocated. Using the three-dimensional spatiotemporal gridding theory, the target low-altitude airspace is discretized in four dimensions: longitude, latitude, altitude, and time, to construct parameterized and semantically defined dynamic airspace units. Step S2: Generate a real-time airspace situation map based on flight plan data, ADS-B surveillance data, numerical weather prediction data, and airspace restriction notice data; and model the target low-altitude airspace allocation problem as a constrained multi-target airspace allocation optimization model. Step S3: Based on the performance requirements analysis of the target low-altitude airspace to be allocated, the target low-altitude airspace is divided into multiple functional layers in the vertical direction, including the take-off and landing and agile operation layer, the conventional cruise layer and the high-speed channel layer. The interval requirements and capacity limits of each functional layer are added to the model established in step S2. Step S4 introduces the functional layer matching metric technique, taking the consistency between the dynamic airspace unit assigned to the aircraft and the expected functional layer corresponding to the aircraft's performance and mission type as the optimization objective, and adding it to the model established in step S2; at the same time, hierarchical admission constraints are added to the constraints, so that the aircraft can only use the dynamic airspace unit of its corresponding authorized functional layer. Step S5: Introduce fuzzy comprehensive evaluation theory, establish a dynamic airspace unit risk rating model, and evaluate and update the risk level of each dynamic airspace unit; based on the risk level of each dynamic airspace unit, minimize the total risk as the optimization objective and add it to the model established in step S2. Step S6: Based on steps S2-S5, an improved genetic algorithm is used to solve the objective function of the constrained multi-target airspace allocation optimization model to obtain the dynamic allocation results of each dynamic airspace unit in the target low-altitude airspace.

[0007] Compared with the prior art, the present invention, employing the above technical solution, has the following technical effects: 1. This invention fundamentally eliminates the airspace isolation and resource idleness problems caused by static airspace delineation by discretizing continuous airspace into parameterized dynamic airspace units (DSUs) and establishing a dynamic mechanism for time-based allocation and idle-time release. Combined with near real-time intelligent allocation using multi-objective optimization algorithms, it can accurately match airspace resources according to rapidly changing operational needs, significantly improving airspace unit turnover rate and overall traffic throughput.

[0008] 2. This invention uses a digital twin environment to simulate and extrapolate allocation schemes, preemptively eliminating potential conflicts; it utilizes real-time monitoring and consistency checks based on Statistical Process Control (SPC) to provide early warnings and rapid intervention for flight deviations; and it optimizes safety parameters through closed-loop feedback. Furthermore, dynamic risk classification management based on fuzzy comprehensive evaluation achieves a shift from a one-size-fits-all approach to precise and differentiated management based on real-time risks, greatly enhancing the system's risk prediction and proactive defense capabilities.

[0009] 3. This invention demonstrates exceptional adaptability to multiple dynamic factors in low-altitude environments, including flight plans, weather conditions, and airspace restrictions. Based on a rolling time-domain optimization framework and real-time data injection, it can quickly respond to emergencies (such as severe weather or urgent missions) and reconstruct airspace allocation schemes in near real-time. This dynamic resilience ensures that the system maintains efficient and orderly operation even under disturbances, exhibiting system resilience unmatched by traditional static models.

[0010] 4. This invention achieves refined management of airspace resources through a three-dimensional management strategy of vertical layering and horizontal grading. Vertically, functional layers are divided according to the operational needs of aircraft with different performance levels, reducing operational interference and improving traffic flow efficiency. Horizontally, differentiated control intensity is implemented based on real-time risks, providing greater flexibility for low-risk operations while ensuring safety. This collaborative management mechanism effectively balances the contradictions between safety, efficiency, and operational freedom.

[0011] 5. This invention breaks through the limitations of the traditional management model, which separates decision-making and evaluation, by constructing a data-driven closed-loop management system. The system can automatically collect operational data, calculate key performance indicators (KPIs), and use machine learning models to analyze the mapping relationship between model parameters and performance, thereby automatically adjusting and optimizing core parameters such as the weight of the optimization target and the safety interval threshold. Attached Figure Description

[0012] Figure 1 This is a flowchart illustrating a dynamic allocation and management method for low-altitude airspace according to the present invention. Figure 2 This is a schematic diagram of the process of constructing parameterized and semantic dynamic spatial units (DSUs) according to the present invention; Figure 3 This is a schematic diagram of the process for establishing a constrained multi-objective spatial allocation optimization model according to the present invention; Figure 4 This is a schematic diagram of the hierarchical airspace management process of the present invention; Figure 5 This is a flowchart illustrating the overall architecture and integration relationship system of this invention. Detailed Implementation

[0013] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0014] like Figure 1 As shown, this invention proposes a dynamic allocation and management method for low-altitude airspace, comprising the following steps: Step S1 involves acquiring basic airspace geographic information, historical and real-time flight track data, an airspace operation rule base, and real-time flight plans and dynamic restriction data. Based on spatiotemporal gridding theory and Geographic Information System (GIS) technology, parameterized and semantic dynamic airspace units (DSUs) are constructed. The unit state vectors and associated attributes of the DSUs are stored in the urban low-altitude airspace management and decision-making platform, serving as a unified data base for the dynamic allocation and management of airspace resources. Employing three-dimensional spatiotemporal gridding theory, the managed airspace is discretized along four dimensions: longitude, latitude, altitude, and time, transforming continuous physical airspace into a discrete, computable spatiotemporal grid.

[0015] Specifically, such as Figure 2 As shown, step S1 includes the following sub-steps: S11: Parametric Modeling Parametric modeling is performed based on the requirements analysis to define the airspace management scope and grid resolution. First, based on a GIS map, the airspace to be managed is clearly defined, and its geographical boundaries are determined. Typically, a city or specific region is used as the boundary, and its latitude and longitude coordinates (lower left corner) are obtained. Top right corner Next, determine the altitude range and define the vertical boundaries of the managed low-altitude airspace.

[0016] S12: Spatial and temporal sampling, including the following sub-steps: S121: Spatial Sampling and Grid Resolution Setting Based on airspace complexity and aircraft performance requirements, and balancing management accuracy with computational complexity, the horizontal grid resolution is dynamically set. Specific strategy: In densely built-up urban areas, a higher resolution (50m × 50m) is used to accurately depict the airspace characteristics around buildings and avoid obstacle conflicts. In suburban or open areas, a lower resolution (200m × 200m) is used to reduce computational load while still meeting basic management needs.

[0017] S122: Vertical Height Layer Division Referring to aviation altitude standards, a vertical resolution of 30 meters per level is set to densely stratify the low-altitude airspace. This division adapts to the diverse altitude operation needs of low-altitude aircraft, ensuring that the vertical management granularity matches the safety separation requirements.

[0018] S123: Time resolution configuration Based on the system's processing capacity and operational dynamism requirements, the time slice length is set as the time resolution: in normal scenarios, the time slice length is 30 seconds, balancing real-time performance and computational efficiency. In highly dynamic environments (such as dense drone take-off and landing areas), a shorter time slice (5 seconds) is used to achieve near real-time control and quickly respond to sudden changes.

[0019] S13: Perform three-dimensional mesh generation for low-altitude airspace. A three-dimensional grid is generated for low-altitude airspace using geographic coordinate transformation and a regular grid generation algorithm. The specific method is as follows: First, a reference coordinate system is selected, using the WGS-84 geographic coordinate system as the spatial reference coordinate system to ensure consistency and standardization of global positioning. Since latitude and longitude are spherical coordinates, directly dividing the grid according to latitude and longitude results in irregular shapes and unequal areas on the plane. Therefore, for the managed area, the spherical latitude and longitude coordinates under the WGS-84 coordinate system need to be transformed to a Cartesian coordinate system using map projection methods (such as UTM projection or Gauss-Kruger projection) to ensure that each grid cell has a regular geometric shape and equal area on the projection plane, thus supporting subsequent equally spaced grid division and distance and area calculations. In short, latitude and longitude coordinates are projected onto a Cartesian reference coordinate system, rather than directly equating latitude and longitude coordinates to WGS-84 spherical coordinates for grid division.

[0020] Mesh generation is performed in the projected Cartesian coordinate system. A regular spatial mesh is generated using an equal-spacing method, and a double-loop algorithm is used to achieve systematic mesh generation. Regarding the relationship between equal-spacing mesh generation and the use of different resolutions in different regions in step S121: Step S121 describes using different mesh resolutions in different regions, with a uniform resolution within the same region, while resolutions can differ between different regions. When generating a regular mesh, the system first performs equal-spacing mesh generation on each sub-region according to its resolution settings. Equal spacing means that the mesh spacing within each sub-region remains consistent, corresponding to the resolution setting of that sub-region, and does not affect the implementation of the regular mesh generation algorithm. A unique three-dimensional spatial index is established for each mesh unit, forming a structured mesh management system.

[0021] S14: Discretized Low-Altitude Airspace Dynamic Time Dimension A discrete-time system model and time-slice round-robin scheduling are constructed to discretize the dynamic time dimension of the low-altitude airspace. First, the system time axis is defined based on Coordinated Universal Time (UTC) to establish a unified system simulation time. Second, time slices are divided using discrete-time system modeling methods, partitioning the continuous time axis into slices of length [missing information]. The time intervals are of equal length. Each interval is a time slice, and they are sequentially numbered to form the basic time unit for discretization management. For example, starting from the beginning time... start, arrive For the first time slice, arrive This is the second time slice, and so on.

[0022] The system advances the simulation clock based on a time-slice round-robin scheduling mechanism, cyclically processing the airspace state according to the time-slice sequence to achieve gradual advancement and periodic updates of airspace management. Finally, it associates the time and space of the low-altitude airspace. For the same spatial location in different time-slice reference frames, the target is considered a completely different and independent management resource unit. Therefore, it is necessary to bind each three-dimensional spatial grid to each time slice, thereby upgrading its unique identifier to a four-dimensional identifier, thus defining an independent and dynamically manageable dynamic airspace unit in the four-dimensional spatiotemporal space.

[0023] S15: Constructing a spatial-temporal grid data structure for low-altitude airspace A low-altitude airspace spatial-temporal grid data structure is constructed using object-oriented programming and a four-dimensional array / database storage method. First, a DSU class / structure is designed, and then a software data structure is designed for each dynamic airspace unit, containing the following static attributes and dynamic states: Static attributes refer to information that is assigned a value once during the system initialization phase based on basic geographic information and the airspace rule base, and remains largely unchanged during system operation. Specifically, these include: ① basic geographic attributes (center coordinates, boundary coordinates, area); ② the static portion of physical constraint attributes (permanent obstacle markers such as buildings, mountains, permanent no-fly zones); ③ social constraint attributes (noise sensitivity index, population density index, privacy protection level – predefined based on geofencing technology). Dynamic status refers to the attributes and status information that change in real time as the system operates. Specifically, it includes: ① Unit status (idle / occupied / reserved / unavailable – updated in real time with flight mission allocation and release); ② Dynamic attributes (passage capability value – dynamically calculated based on real-time wake separation and radio interference index (RFI); risk level – dynamically assessed by a fuzzy comprehensive evaluation model); ③ The dynamic part of physical constraint attributes (temporary obstacle markings such as construction cranes, temporary no-fly zones such as security areas for major events).

[0024] S16: Initialize and dynamically update the low-altitude airspace spatial-temporal grid Rule-based reasoning and real-time data injection are used to initialize and dynamically update the low-altitude airspace spatial-temporal grid model. At system startup, static attributes of all DSUs are assigned values ​​based on fundamental geographic information and the airspace rule base for static initialization. For example, the status of DSUs containing buildings or permanent no-fly zones is initialized to "unavailable." Subsequently, an update loop is established to maintain dynamic updates of the system after it begins operation.

[0025] In each time slice Then, the system clock increments by 1, generating a new batch of future DSUs with an "idle" status. Furthermore, event-driven mechanisms ensure the system receives external data (flight plans, weather data, surveillance data, etc.) in real time, triggering dynamic resource allocation or status update processes. When a new flight plan is executed, the system attempts to allocate a channel for it in the future time slice's DSU sequence; when surveillance data indicates that an aircraft has left a DSU, the DSU's status is immediately updated to "idle," achieving real-time resource recovery and reuse.

[0026] Step S2: Establish a constrained multi-objective spatial allocation optimization model The system accesses flight plans (4D tracks), ADS-B and other monitoring data, numerical weather prediction (NWP) data, and airspace restriction notices (NOTAM) via a data bus. Inconsistent and uncertain information is processed using data fusion algorithms to generate a unified real-time airspace situation map. Subsequently, a low-altitude airspace dynamic allocation optimization model is constructed, modeling the airspace allocation problem as a constrained multi-objective optimization problem, such as... Figure 3 As shown.

[0027] S21: Establish the objective function Since this model needs to optimize multiple objectives simultaneously, a weighted summation method or Pareto optimality method is usually used. Here, the weighted summation method is used to transform the multi-objective problem into a single objective, but it is important to note that the weight settings need to balance the importance of each objective.

[0028] Objective 1: Maximize airspace utilization , The above formula means using as many spatial units as possible. Indicates airspace utilization rate. For binary decision variables, , Number the flight mission. For flight mission collection, Numbering of dynamic spatial domain units, For a dynamic spatial domain unit set, Number the time slices. , The number of time slices to be allocated for the time period. A set of time slices; Indicates flight mission In time slice Assigned to use dynamic spatial unit , Indicates flight mission In time slice Dynamic spatial unit not assigned .

[0029] Objective 2: Minimize total conflict risk Conflict risk can be measured by the probability of conflict between flight missions. Define the conflict risk function. Indicates flight mission and In time slice Simultaneously using dynamic spatial units Given the conflict risk, the objective function is: , in, This indicates the overall risk of conflict.

[0030] Objective 3: Minimize total track deviation: set up Indicates flight mission Using dynamic spatial units Compared to flight missions If the deviation from the original planned trajectory is true, then the objective function is: , in, This indicates a deviation from the total flight path.

[0031] Objective 4: Maximize task priority: set up Indicates flight mission The priority is determined by the numerical value; higher priority tasks are allocated more spatial units than required. This can be expressed as maximizing priority-weighted allocation: , in, Indicates task priority.

[0032] S22: Set constraints Constraint 1: Spatial Unit Availability Constraint Each airspace unit can be used by a maximum of one flight mission in the same time slice to avoid conflicts: .

[0033] Constraint 2: Flight Mission Continuity Constraint Each flight mission must be assigned to an airspace unit in adjacent time slices, and these airspace units must be spatially adjacent to ensure track continuity. Let... Representation and dynamic spatial unit Adjacent dynamic spatial units The set, then for each flight mission and each time slice (Except for the last time slice), there are: , , in, This indicates that, except for the last time slice The set of remaining time slices; the first constraint ensures that each task is assigned a spatial unit in each time slice, and the second constraint ensures that the spatial units of adjacent time slices are adjacent.

[0034] Constraint 3: Aircraft Performance Constraints The maximum turning radius and rate of climb limits of an aircraft can be reflected through the accessibility between airspace units. Considering the aircraft's maximum speed, turning radius, etc., the following definitions are established. Indicates from dynamic spatial unit It can be reached in the next time slice. The constraint can then be written as: , if from If it is unattainable, then If so, the constraint will prohibit allocation.

[0035] Constraint 4: Safety Interval Constraint In addition to conflict risk objectives, safety intervals are ensured through hard constraints. Using the REICH model, it is required that no two flight missions can simultaneously use adjacent airspace units (including their own units) in any time slice. , This constraint ensures that there is at least one airspace unit between flight missions.

[0036] Constraint 5: Takeoff and landing constraints When a flight mission submits a flight plan, the plan clearly includes the takeoff position (such as specific takeoff and landing field coordinates) and the destination position (such as target landing field coordinates). The system maps the takeoff position to the corresponding initial airspace unit. Map the destination location to the corresponding target airspace unit. .therefore, and The value is already determined before the optimization model is solved, so there is no need to make a decision through the optimization process.

[0037] Step S3: Based on the aforementioned established Dynamic Spatial Unit (DSU) and constrained multi-objective spatial allocation optimization model, the spatial domain is managed hierarchically and hierarchically to achieve refined control, such as... Figure 4 As shown, it includes the following sub-steps: S31: Divide the airspace into multiple functional layers First, a hierarchical management strategy based on fuzzy comprehensive evaluation is formulated in the vertical dimension. Based on airspace performance requirements analysis, the airspace is divided into three functional layers in the vertical direction. The specific division and parameters of the three functional layers are as follows: Takeoff and landing and agile operation layer (applicable to speeds below 50 knots, with a smaller minimum horizontal / vertical separation due to the slower aircraft speed); regular cruise layer (applicable to speeds of 50-100 knots, with a medium minimum horizontal / vertical separation); high-speed passage layer (applicable to speeds above 100 knots, with a larger minimum horizontal / vertical separation due to the greater safety margin required at higher speeds).

[0038] S32: Hierarchical parameterized configuration Airspace performance standard setting technology is applied to each functional layer to configure differentiated DSU basic parameters. The basic parameters include: a. minimum horizontal spacing, i.e. the minimum horizontal safe distance between aircraft within the layer; b. minimum vertical spacing, i.e. the minimum vertical safe distance between aircraft within the layer; c. maximum capacity of the layer, i.e. the maximum number of aircraft that each DSU can accommodate at the same time in this functional layer; d. access aircraft performance level, i.e. the performance (speed, maneuverability, etc.) requirements for aircraft allowed to enter this layer.

[0039] S33: Layer Capacity Differentiation Calculation Based on the interval standards set for each functional layer, the capacity of each functional layer's DSU (Dynamic Airspace Unit) is dynamically calculated using a capacity calculation model, which is the number of aircraft that can be accommodated simultaneously.

[0040] S34: Optimization Model Constraint Integration By using hierarchical constraint mapping technology, specific parameters (interval requirements, capacity limits) of each functional layer are integrated into the spatial allocation optimization model in step S2 as key constraints.

[0041] Step S4: Set up a priority allocation mechanism for functional layers In the optimization model of step S2, when allocating airspace units for flight missions, the aircraft will be preferentially allocated to the corresponding functional layer based on the performance (cruise speed, etc.) and mission type (take-off and landing, cruise) declared by the aircraft.

[0042] S41: Functional Layer Matching Degree Target Construction In the objective function of the optimization model, a functional layer matching metric is introduced to increase the consistency between the DSUs assigned to the aircraft and the expected functional layers corresponding to its performance and mission type as an optimization objective.

[0043] S42: Hierarchical Admission Constraint Settings Based on aircraft performance certification and mission type, the hierarchical admission rule engine stipulates in the constraints of the optimization model that the aircraft can only use the DSU of its corresponding authorized function layer.

[0044] S43: Priority Strategy Integration Optimization The matching degree objective and hierarchical admission constraints mentioned above are integrated into the spatial allocation optimization model in step S2 through a multi-objective constraint integration method, guiding the solver to prioritize the generation of allocation schemes that conform to the original intention of functional layer division.

[0045] Step S5: Develop a hierarchical management strategy based on fuzzy comprehensive evaluation. A hierarchical management strategy based on fuzzy comprehensive evaluation is formulated along the horizontal / risk dimension. Fuzzy comprehensive evaluation theory is introduced to establish a dynamic risk rating model for airspace units. This includes the following sub-steps: S51: Construction of Risk Assessment Indicator System Using multi-source information fusion technology, the following indicators are selected to evaluate the risk of each DSU: real-time traffic flow density, the number of aircraft in the DSU and its neighboring DSUs in the current time slice; weather conditions, including visibility, wind speed, etc.; distance to obstacles, the distance between the DSU and the nearest obstacle or building; and monitoring coverage, the degree to which the DSU is covered by monitoring equipment (radar, ADS-B receiving station).

[0046] S52: Determining Indicator Weights Using the Analytic Hierarchy Process (AHP) or a fuzzy inference system, the relative weights of each risk assessment indicator are determined through expert evaluation and consistency testing, forming a weight vector.

[0047] S53: Implementation of Fuzzy Comprehensive Evaluation Based on fuzzy mathematics theory, a membership function and a fuzzy rule base are established. The data of each indicator of each DSU are fuzzified, and its comprehensive risk value is calculated through fuzzy synthesis operation.

[0048] S54: Dynamic Classification of Risk Levels Based on the comprehensive risk value, DSU is dynamically divided into three levels—relaxed (low risk), standard (medium risk), and strict (high risk)—using threshold classification technology, providing a basis for differentiated management.

[0049] S55: Establish membership function For each factor, it is necessary to determine its degree of belonging to each risk level. For traffic flow density, the following definitions apply: density below 0.1 is low risk (unit: number of aircraft / DSU area); density between 0.1 and 0.3 is medium risk; and density above 0.3 is high risk. Then, the actual measured values ​​are converted into membership degrees to the three risk levels using a membership function.

[0050] Based on fuzzy set theory, a triangular or trapezoidal membership function is designed for each risk factor to quantify the degree to which the actual measured value belongs to each risk level.

[0051] S56: Implementation of fuzzification Using fuzzy processing technology, the actual measured values ​​of each factor are input into the membership function, which outputs the membership vectors of each factor to three risk levels (low, medium, and high). The parameters of the membership function are calibrated and optimized using a parameter tuning algorithm to ensure that it accurately reflects the risk change patterns in the actual operating environment.

[0052] S57: Results of Synthetic Fuzzy Comprehensive Evaluation Using weight vectors and membership matrices, fuzzy synthesis operations are performed to obtain the membership vector of each DSU with respect to the comment set V. Commonly used synthesis operators include the M(∧,∨) operator and the weighted average operator. Finally, based on the membership vectors, the risk level of the DSU is determined according to the maximum membership principle or the weighted scoring method. The determination results are then transformed into a system-recognizable risk level identifier using standardized coding techniques, providing standardized input for subsequent hierarchical management.

[0053] S58: Risk Level Dynamic Update Because risk assessment metrics change in real time, the risk level of each DSU needs to be updated dynamically. An update cycle can be set, or updates can be made when a metric's change exceeds a certain threshold.

[0054] Through real-time processing technology of multi-source data streams, we continuously monitor various dynamic indicators that affect the risk level of DSU (such as traffic flow, meteorological data, monitoring coverage, etc.).

[0055] A dual-mode triggering mechanism is adopted, combining timer periodic triggering and event triggering based on threshold comparison algorithm to intelligently determine when to start the risk reassessment process.

[0056] When the triggering conditions are met, the fuzzy comprehensive evaluation engine is invoked to recalculate the comprehensive risk value and membership vector of the DSU based on the latest indicator data.

[0057] The risk level attributes of DSUs are updated through the status management service, and the change notifications are distributed to all relevant subsystems using a publish-subscribe mechanism to ensure timely adjustments to control strategies.

[0058] S59: Add risk level attribute A risk level attribute is added to each DSU, which is dynamically calculated by the fuzzy comprehensive evaluation model in this step. In the optimization model of step S2, the risk level will affect the allocation of airspace units. For example, high-risk DSUs may be restricted in use, or only available for high-priority flight missions. This can be achieved by introducing the risk level into the objective function or constraints of the optimization model. In the objective function, an objective term can be added, namely minimizing the total risk, expressed as: , in, Indicates total risk. For dynamic spatial units Risk factor, dynamic spatial unit When it is low risk, Dynamic spatial unit When it is medium risk, Dynamic spatial unit When it is high risk, At the same time, constraints are set, such as prohibiting the use of DSUs with excessively high risk levels for flight missions.

[0059] S510: Policy Mapping Different levels of control will be implemented based on the risk level of the DSU. The risk levels are set into three levels: Relaxed (Low Risk): Autonomous flight with a reporting system is implemented; the aircraft can autonomously plan its trajectory and only needs to report. Standard (Medium Risk): Flight must follow precise instructions based on the consensus trajectory (TBO); the aircraft must fly according to the assigned 4D trajectory. Strict (High Risk): TBO must be followed, and enhanced monitoring is required, requiring the aircraft to be equipped with and activate ADS-B Out, or to be continuously tracked by radar. These strategies will be reflected in the flight management rules and, after the optimization model is solved in step 2, will be issued to the aircraft as flight commands.

[0060] By using constraint mapping technology, the hierarchical control rules are integrated as key constraints into the optimization model of step S2 to ensure that the generated allocation scheme meets the control requirements of each risk level.

[0061] Step S6: Solving the algorithm Since the optimization model in step S2 is a large-scale mixed-integer programming problem, and an NP-hard problem, a metaheuristic algorithm is used to solve it. First, a genetic algorithm is designed, using a three-dimensional matrix encoding, where each individual represents an allocation scheme, i.e., a... A three-dimensional binary matrix with elements of Next, an initial population is generated, and individuals satisfying constraints 1 and 2 are randomly generated, ensuring that each task has exactly one spatial unit in each time slice, and that spatial units in adjacent time slices are adjacent. A fitness function, i.e., a comprehensive objective function, is then established. Next, genetic operations are performed, using roulette wheel selection or tournament selection methods. Due to constraint 2 (continuity), simple crossover cannot be performed; therefore, a crossover operation that maintains continuity is designed, i.e., randomly selecting a task and its time slice for exchange. Finally, the spatial unit of a task in a certain time slice is randomly changed, but it must be ensured that it is adjacent to the spatial units of the preceding and following time slices, thus completing the design of the genetic algorithm. This includes the following sub-steps: S61: Chromosome Coding Design Using three-dimensional matrix encoding technology, each individual (allocation scheme) is represented as a... A three-dimensional binary matrix, in which elements Identifying flight missions In time slice Does it occupy an airspace unit? .

[0062] S62: Feasible solution for population initialization An initial population is generated by satisfying constraints through an initialization method, ensuring that each individual satisfies the constraint that each task occupies only one unit per time slice and that adjacent time slice units are adjacent.

[0063] S63: Establishment of Fitness Evaluation System Construct a multi-objective fitness function, which integrates the objective function. As an individual evaluation standard, it guides the direction of population evolution.

[0064] S64: Design of Genetic Operators (a) Selection Operation Using roulette wheel selection or tournament selection mechanisms, based on individual fitness values ​​(comprehensive objective function) Select high-quality individuals from the current population to enter the mating pool. Individuals with higher fitness have a greater probability of being selected.

[0065] (ii) Special crossover operations – crossover methods to maintain track continuity Since the airspace units allocated to each flight mission in adjacent time slices must be spatially adjacent (Constraint 2: Flight Mission Continuity), traditional uniform or single-point crossings would disrupt track continuity, rendering offspring infeasible. Therefore, the following crossing operation is designed specifically to maintain track continuity: A. Randomly select two parent individuals from the mating pool. and Each individual is A three-dimensional binary matrix; from the flight mission set Randomly select a flight mission As an intersection object; B. Randomly determine a continuous time segment. ( This range covers flight missions. Part or all of the flight time; C. Parental individuals Mid-flight mission In time segment All allocation schemes within (i.e.) Replace the entire parent individual with the parent individual. Mid-flight mission The allocation scheme within the same time frame. Simultaneously, the parent individuals... Mid-flight mission exist The internal allocation scheme was replaced with The original plan remains unchanged. The allocation of remaining flight missions and time slices remains the same. D. Verify the feasibility of the offspring generated by the crossover operation: check at the exchange boundary ( and Between and and (between), flight mission Does the trajectory still satisfy the spatial adjacency constraint? If not, then locally adjust 1-2 adjacent dynamic spatial units at the boundary along the trajectory direction to restore trajectory continuity. The adjustment method is as follows: from the set of adjacent dynamic spatial units in the boundary time slice, select those that satisfy the comprehensive objective function. The dynamic spatial unit with the smallest descent is used as the transition unit.

[0066] (III) Special Mutation Operations – Neighborhood Preservation Mutation Methods Traditional mutation operations randomly flip a certain However, this would violate constraint 1 (there is exactly one spatial unit per time slice) and constraint 2 (track continuity). Therefore, the following mutation operation is designed specifically to preserve neighborhood relationships: E. Select any one of the offspring individuals generated by the crossover operation. ,exist Randomly select a flight mission and a time slice ( ); F. Obtaining Flight Missions In time slice Allocated airspace units (Right now ), and the dynamic spatial units allocated to the time slices before and after it. (Time slice) )and (Time slice) ); G. Determine the set of candidate variant spatial units, including the candidate variant spatial units in this set. All three of the following conditions must be met: Condition 1: Cell availability, candidate mutated spatial cells in time slice Not occupied by other flight missions; Condition 2: Forward adjacent, from Departure, in a time slice Reachable ,Right now (when hour); Condition 3: Backward adjacent, from Departure, in a time slice Reachable ,Right now (when hour); H. If the candidate mutation airspace unit set is empty (i.e., there are no candidate units that simultaneously meet the above conditions), then this mutation is abandoned, and a new flight mission and time slice are selected; if it is not empty, then a candidate mutation airspace unit is randomly selected from the candidate mutation airspace unit set. Flight mission In time slice The allocation from Change to , that is to say , ;because If the forward and backward adjacency conditions are met, the modified track will automatically maintain continuity. I. Perform the above mutation operation on each offspring individual with a preset mutation probability.

[0067] (iv) Crossover and mutation probabilities Crossover probability is typically set between 0.6 and 0.9, and mutation probability is typically set between 0.01 and 0.1. Specific values ​​need to be determined through experimental optimization based on the problem size and real-time requirements.

[0068] S65: Iterative Optimization and Convergence By iteratively evolving through generations, the quality of the solution is gradually improved, and a satisfactory solution that meets the requirements of real-time computing is finally obtained.

[0069] S66: Linearization Since the conflict risk term in the objective function is a quadratic term, it can be transformed into a linear term using linearization techniques. This can be achieved by introducing auxiliary variables. express (i.e., two different flight missions) and In time slice Simultaneously occupying airspace units The indicator variables are defined, and corresponding linear inequality constraints are added. After the above linearization process, the spatial dynamic allocation optimization model (including the linearized objective function and all linear constraints) is transformed into a standard mixed-integer linear programming (MILP) model. For small-scale problem instances, commercial or open-source MILP solvers (such as CPLEX, Gurobi, SCIP, etc.) can be used for exact solutions.

[0070] S67: Establish a real-time operation mechanism In actual operation, a rolling horizontal optimization strategy is adopted: Based on the current airspace status and known flight plans, optimize airspace allocation for the future period; execute allocation instructions for the current time slice; as time progresses, add new flight requests and updated information, and re-optimize the allocation for the next time slice.

[0071] The above strategy decomposes large-scale problems into multiple consecutive smaller-scale problems, satisfying real-time requirements. Based on the above steps, a dynamic spatial allocation mechanism based on multi-objective optimization algorithms can be established. Due to the complexity of the problem, the model and parameters need to be adjusted according to the specific scenario in practical applications.

[0072] like Figure 5 As shown, this invention also establishes the overall system architecture and integration relationships. This closed-loop management system, acting as the brain and learning center of the entire low-altitude dynamic airspace management framework, forms a deeply coupled organic whole with the first three steps. The system constructs a virtual mapping through digital twin technology, ensures operational safety through real-time monitoring, drives system optimization through performance evaluation, and ultimately forms a complete data-driven closed loop of perception-decision-execution-evaluation-optimization.

[0073] 1) Establish a deep integration mechanism with previous steps. The closed-loop management system is directly built upon the dynamic spatial unit constructed in step S1. Each DSU has a corresponding virtual entity in the digital twin environment, carrying the real-time status, historical trajectory, and performance indicators of the spatial unit. The system uses the parameterized attributes of the DSU as the basic data source for monitoring and evaluation, while simultaneously feeding the optimization results back to the DSU's status update. The multi-objective optimization algorithm in step S6 runs as the core decision engine in the closed-loop system. The system dynamically adjusts the constraints and objective weights of the optimization model through real-time monitoring data, while using the digital twin environment to pre-validate the optimization results. The allocation scheme generated by the optimization algorithm must undergo rigorous simulation testing in the twin environment before execution to ensure its safety and feasibility. The risk level and management strategy determined in step S5 serve as dynamic constraints in the closed-loop system. The system adjusts the monitoring intensity and control rules based on real-time risk assessment, while simultaneously feeding the actual operating data back to the risk assessment model to achieve adaptive optimization of the risk judgment criteria. Different risk levels correspond to different monitoring frequencies and evaluation standards for different areas.

[0074] 2) Establish a digital twin environment and verification mechanism The construction of the digital twin environment is based on high-precision 3D modeling technology and a real-time data synchronization mechanism. First, a digital copy of the physical airspace is constructed using Geographic Information System (GIS) and Building Information Modeling (BIM) to ensure geometrical consistency between the virtual environment and the real world. Second, a multi-source data access bus is established to inject real-time data on aircraft status, weather conditions, and airspace restrictions into the twin environment, maintaining synchronization between the virtual and physical states. The synchronization mechanism adopts a publish-subscribe model, with various data sources acting as publishers pushing updates to the data bus, and components within the twin environment acting as subscribers consuming data on demand. To ensure real-time performance, the system sets differentiated update frequencies and latency requirements for different types of data; for example, monitoring data requires sub-second updates, while meteorological data can tolerate delays of several minutes.

[0075] 3) Simulation model derivation and safety verification During the simulation phase before command issuance, the system employs a multi-resolution modeling approach. For routine verification, a lightweight, fast simulation model is used to complete preliminary safety checks within seconds. For complex scenarios or high-risk operations, a high-fidelity physical simulation model is used to accurately simulate aircraft dynamics and environmental interactions. Safety verification covers three dimensions: conflict safety verification predicts potential flight conflicts in allocation schemes through probabilistic calculations based on an improved REICH model; performance safety verification evaluates the impact of the scheme on overall system performance, including airspace utilization, average delay, and other indicators; and operational safety verification considers the impact of environmental factors such as weather conditions and equipment status on the feasibility of the scheme. During verification, the system pays special attention to edge cases and extreme scenarios, randomly generating disturbance factors using the Monte Carlo method to test the robustness of the allocation scheme. Only schemes that pass comprehensive verification are approved for execution; otherwise, they are returned to the optimization engine for recalculation.

[0076] 4) Trajectory monitoring based on statistical process control The real-time monitoring system employs statistical process control methods to continuously monitor the consistency between the aircraft's actual trajectory and assigned commands. The system establishes individual control charts and moving range control charts for each flight mission, calculating the statistical characteristics of position and time deviations in real time. The centerline of the control chart is dynamically adjusted based on historical performance data, while control limits are adaptively calculated according to real-time traffic density and airspace complexity. When deviation data points exceed control limits, or when a non-random pattern such as seven consecutive increases / decreases occurs, the system identifies this as a statistical anomaly and triggers a tiered alarm mechanism. Consistency assessment not only focuses on absolute deviation values ​​but also considers the trend of deviation development. The system smooths the deviation sequence using an exponentially weighted moving average method, identifying potential deviation trends in advance and shifting from passive alarms to proactive warnings.

[0077] 5) Intelligent Conflict Detection and Resolution The conflict detection algorithm employs a multi-level design. Primary detection uses simple geometric relationships for rapid screening; intermediate detection considers aircraft performance constraints and intent information; and advanced detection incorporates uncertainty modeling to assess conflict probability rather than simply a binary "conflict" status. When a potential conflict is detected, the system selects an appropriate resolution strategy based on the conflict type, severity, and available response time. For minor conflicts, strategies with minimal operational impact, such as speed adjustments, are prioritized; for urgent conflicts, multiple methods, such as altitude changes and lateral drift, may be combined. Conflict resolution decisions consider overall airspace effectiveness, avoiding local optimization that could lead to global performance degradation. The system rapidly simulates and evaluates the impact of various resolution schemes on surrounding airspace, selecting the scheme with the lowest system cost. Simultaneously, it provides decision support to human controllers, displaying a comparison of the advantages and disadvantages of each scheme and a recommendation level.

[0078] 6) Establish a multi-dimensional performance indicator system The system has established a comprehensive key performance indicator (KPI) system to evaluate system performance from multiple dimensions, including efficiency, safety, capacity, and reliability. Efficiency indicators include airspace unit turnaround rate, average delay time, and fuel efficiency; safety indicators cover conflict alarm frequency, safety interval violation rate, and risk exposure time; capacity indicators focus on peak throughput and saturation distribution; and reliability indicators assess system availability and robustness. All indicators are standardized to eliminate the influence of dimensions, and their relative weights are determined using the analytic hierarchy process (AHP). The system periodically calculates a comprehensive performance score, providing a unified benchmark for performance comparisons across different periods. Furthermore, for different types of operational scenarios (such as intensive takeoffs and landings, high-speed cruise, etc.), the system adjusts the indicator weights to highlight the key performance dimensions under those scenarios.

[0079] 7) Data-driven parameter optimization The system collects a large amount of operational data, including environmental conditions, decision parameters, and resulting performance metrics, to construct a dataset mapping parameter-performance relationships. A random forest regression model is used to analyze the impact of each decision parameter on system performance. The advantage of random forests lies in their ability to handle nonlinear relationships and interaction effects between parameters, while providing feature importance ranking to identify the key parameters with the greatest impact on performance. The optimization process employs Bayesian optimization, efficiently searching for the optimal configuration in the parameter space while ensuring a balance between exploration and utilization. For continuous parameters, a Gaussian process is used to model the objective function; for discrete parameters, a tree-structured Parzen estimator is employed. The optimization objective comprehensively considers both short-term performance and long-term stability, avoiding overfitting to specific scenarios.

[0080] 8) Algorithm iteration based on field validation The system undergoes algorithm verification and optimization in controlled environments such as unmanned aerial vehicle (UAV) test application bases. Field verification employs a progressive strategy: first, the correctness of the core algorithm is verified in simple scenarios; then, the system's robustness is tested in complex scenarios; and finally, the system's extreme performance is evaluated in high-density scenarios. During verification, various disturbances, such as communication delays, sensor errors, and sudden weather changes, are intentionally introduced to test the system's performance under non-ideal conditions. Simultaneously, edge cases are artificially designed to verify the system's safety and reliability in handling abnormal situations. Based on the field verification results, the system performs targeted optimizations to the route planning algorithm. Optimization directions include: improving the cost function to better balance multiple optimization objectives, enhancing the planning algorithm's adaptability to uncertainty, and improving computational efficiency under large-scale problems. Each algorithm iteration undergoes regression testing to ensure that the new version does not experience performance regressions in existing functions.

[0081] 9) Set the adaptive control loop The system implements a complete adaptive control loop, with each loop comprising five stages: perception, decision-making, execution, evaluation, and optimization. The loop frequency is dynamically adjusted based on system load, maintaining a second-level cycle under normal conditions and appropriately extending the cycle under high load to ensure decision quality. Data transfer and state synchronization between loop stages are achieved through shared memory and message queues. The system maintains a global data context to ensure that each stage operates based on a consistent system state. Simultaneously, a transaction mechanism ensures the atomicity and consistency of state updates, preventing inconsistencies in system state due to partial failures. The control loop possesses degradation and fault tolerance capabilities. When a component fails, the system can automatically switch to a backup algorithm or a simplified process to ensure the continuous operation of basic functions. For example, when the optimization algorithm times out, the system can switch to a rule-based emergency decision-making mode.

[0082] 10) Accumulate operational experience and domain knowledge The continuous learning and knowledge accumulation system constantly accumulates operational experience and domain knowledge through a continuous learning mechanism. The knowledge base stores successful cases and lessons learned from failures in various scenarios, providing references for subsequent decision-making. Cases are represented in a standardized manner, including dimensions such as scenario characteristics, decision parameters, execution results, and performance evaluation. The case-based reasoning system can quickly retrieve relevant cases and learn from historical experience when encountering similar scenarios. Simultaneously, the system periodically performs cluster analysis on the case library to identify typical operating modes and the optimal decision-making strategies under each mode. The system also establishes an online learning mechanism that can fine-tune model parameters based on real-time feedback. The learning process uses an incremental update method to avoid the computational overhead of global retraining. Furthermore, a sliding window mechanism ensures that the system can adapt to gradual changes in the operating environment and is not misled by short-term fluctuations. This complete closed-loop management system, through the organic combination of digital twins, real-time monitoring, performance evaluation, and continuous optimization, ensures the safety, efficiency, and adaptability of the low-altitude dynamic airspace management system in actual operation, providing reliable technical support for the large-scale commercial application of urban air traffic.

[0083] Based on the same inventive concept, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the aforementioned dynamic allocation and management method for low-altitude airspace.

[0084] Based on the same inventive concept, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned dynamic allocation and management method for low-altitude airspace.

[0085] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0086] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0087] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0088] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0089] The above embodiments are merely illustrative of the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solutions based on the technical concept proposed in this invention shall fall within the scope of protection of this invention.

Claims

1. A dynamic allocation and management method for low-altitude airspace, characterized in that, Includes the following steps: Step S1: Determine the target low-altitude airspace range and time period to be allocated. Using the three-dimensional spatiotemporal gridding theory, the target low-altitude airspace is discretized in four dimensions: longitude, latitude, altitude, and time, to construct parameterized and semantically defined dynamic airspace units. Step S2: Generate a real-time airspace situation map based on flight plan data, ADS-B surveillance data, numerical weather prediction data, and airspace restriction notice data; and model the target low-altitude airspace allocation problem as a constrained multi-target airspace allocation optimization model. Step S3: Based on the performance requirements analysis of the target low-altitude airspace to be allocated, the target low-altitude airspace is divided into multiple functional layers in the vertical direction, including the take-off and landing and agile operation layer, the conventional cruise layer and the high-speed channel layer. The interval requirements and capacity limits of each functional layer are added to the model established in step S2. Step S4 introduces the functional layer matching metric technique, taking the consistency between the dynamic airspace unit assigned to the aircraft and the expected functional layer corresponding to the aircraft's performance and mission type as the optimization objective, and adding it to the model established in step S2; at the same time, hierarchical admission constraints are added to the constraints, so that the aircraft can only use the dynamic airspace unit of its corresponding authorized functional layer. Step S5: Introduce fuzzy comprehensive evaluation theory, establish a dynamic airspace unit risk rating model, and evaluate and update the risk level of each dynamic airspace unit; based on the risk level of each dynamic airspace unit, minimize the total risk as the optimization objective and add it to the model established in step S2. Step S6: Based on steps S2-S5, an improved genetic algorithm is used to solve the objective function of the constrained multi-target airspace allocation optimization model to obtain the dynamic allocation results of each dynamic airspace unit in the target low-altitude airspace.

2. The dynamic allocation and management method for low-altitude airspace according to claim 1, characterized in that, The specific process of step S1 is as follows: Step S11: Determine the range of the target low-altitude airspace to be assigned, set the horizontal grid resolution and vertical resolution, and use geographic coordinate transformation and regular grid generation algorithm to divide the target low-altitude airspace into three-dimensional spatial grids. Step S12: Determine the time period to be allocated, set the time resolution, divide the time period to be allocated into several time slices according to the time resolution, and number the time slices in sequence. Step S13: Using spatiotemporal correlation technology, each three-dimensional spatial grid is bound to each time slice to construct a dynamic spatial unit. The dynamic spatial unit includes static attributes and dynamic states. The static attributes include: basic geographic attributes, permanent obstacle markers and social constraint attributes. The dynamic states include: unit state, real-time accessibility, dynamic risk level and temporary obstacle markers.

3. The dynamic allocation and management method for low-altitude airspace according to claim 2, characterized in that, The specific process of step S11 is as follows: Determine the geographical and vertical boundaries of the target low-altitude airspace that needs to be managed; The horizontal grid resolution is set as follows: 50m×50m resolution in densely populated urban areas, and 200m×200m resolution in suburban or open areas. The vertical resolution is set to 30 meters per level. Using the WGS-84 geographic coordinate system as the spatial reference, the spherical latitude and longitude coordinates under the WGS-84 geographic coordinate system are transformed to a plane rectangular reference coordinate system through map projection methods to generate plane rectangular coordinates; in the projected plane rectangular coordinate system, the grid is divided according to the set horizontal grid resolution. In the vertical direction, the target's low-altitude airspace is layered according to vertical resolution to obtain a three-dimensional spatial grid.

4. The dynamic allocation and management method for low-altitude airspace according to claim 2, characterized in that, In step S2, the objective function of the constrained multi-objective spatial allocation optimization model includes: Maximize airspace utilization: ， in, Indicates airspace utilization rate. Number the flight mission. For flight mission collection, Numbering of dynamic spatial domain units, For a dynamic spatial domain unit set, Number the time slices. , The number of time slices to be allocated for the time period. A set of time slices For binary decision variables, , Indicates flight mission In time slice Assigned to use dynamic spatial unit , Indicates flight mission In time slice Dynamic spatial unit not assigned ; Minimize total conflict risk: ， in, Indicates the overall risk of conflict; Let be the conflict risk function, representing the flight mission. and In time slice Simultaneously using dynamic spatial units Conflict risk; For binary decision variables, ; Minimize total track deviation: ， in, This indicates a deviation from the total flight path. Indicates flight mission Using dynamic spatial units Compared to flight missions Deviation from the original planned trajectory; Maximize task priority: ， in, Indicates task priority. Indicates flight mission Priority.

5. The dynamic allocation and management method for low-altitude airspace according to claim 4, characterized in that, In step S2, the constraints of the constrained multi-objective spatial allocation optimization model include: Constraint 1: Spatial cell availability constraint Each dynamic airspace unit can be used by a maximum of one flight mission in the same time slice: ， Constraint 2: Flight Mission Continuity Constraint ， ， in, Representation and dynamic spatial unit Adjacent dynamic spatial units The set, This indicates that, except for the last time slice The set of remaining time slices; the first formula ensures that each task is assigned a dynamic spatial unit in each time slice, and the second formula ensures that the dynamic spatial units of adjacent time slices are adjacent; Constraint 3: Aircraft performance constraints ， in, Indicates from dynamic spatial unit In time slice Reachable If from Unreachable ,but This means prohibiting the allocation; Constraint 4: Safety Interval Constraint No two flight missions can use adjacent dynamic airspace units simultaneously in any time slice: ， Constraint 5: Takeoff and Landing Constraints Takeoff and landing constraints require each flight mission to begin from the initial dynamic airspace unit and end at the target dynamic airspace unit, which are determined by the takeoff position and destination position specified in the flight plan, respectively.

6. The dynamic allocation and management method for low-altitude airspace according to claim 1, characterized in that, In step S3, the takeoff and landing and agile operation layer is suitable for aircraft takeoff, landing and low-altitude agile operations, with aircraft speed below 50 knots; the conventional cruise layer is suitable for aircraft conventional cruise flight, with aircraft speed between 50 and 100 knots; and the high-speed passage layer is suitable for aircraft high-speed flight, with aircraft speed above 100 knots. Spacing requirements are set for each functional layer, including: minimum horizontal spacing and minimum vertical spacing. The minimum horizontal spacing is the minimum horizontal safe distance between aircraft within the layer, and the minimum vertical spacing is the minimum vertical safe distance between aircraft within the layer. Based on the interval requirements set for each functional layer, a capacity calculation model is used to calculate the maximum number of aircraft that a dynamic airspace unit within each functional layer can simultaneously accommodate.

7. The dynamic allocation and management method for low-altitude airspace according to claim 1, characterized in that, In step S5, traffic flow density, weather conditions, distance to obstacles, and monitoring coverage are selected as risk assessment indicators for each dynamic airspace unit; three risk levels are set: low risk, medium risk, and high risk. The relative weights of each risk assessment indicator are determined using the analytic hierarchy process (AHP) to form a weight vector. Design a triangular or trapezoidal membership function for each risk assessment indicator to quantify the degree to which the actual measured value belongs to each risk level; Input the actual measured values ​​of each risk assessment indicator into the membership function, and output the membership matrix of each risk assessment indicator to the three risk levels; Using weight vectors and membership matrices, fuzzy synthesis operations are used to obtain the membership vectors of each dynamic spatial unit to the three risk levels, thereby determining the risk level to which each dynamic spatial unit belongs. Set an update cycle to dynamically update the risk level of each dynamic airspace unit; Based on the risk level of each dynamic spatial unit, minimizing the total risk is taken as the optimization objective and added to the model established in step S2; the formula for minimizing the total risk is as follows: ， in, Indicates total risk. For binary decision variables, , For dynamic spatial units Risk factor, dynamic spatial unit When it is low risk, Dynamic spatial unit When it is medium risk, Dynamic spatial unit When it is high risk, ; Different levels of control are implemented based on the risk level of the dynamic airspace unit: when the risk is low, the control intensity is relaxed, autonomous flight is implemented with a reporting system, and the aircraft plans its own trajectory; when the risk is medium, the control intensity is standard, and the aircraft follows the instructions to fly based on a consensus trajectory; when the risk is high, the control intensity is strict, the aircraft implements the instructions to fly based on a consensus trajectory under enhanced surveillance, and the aircraft is required to be equipped with and activate ADS-B surveillance equipment or radar for continuous tracking.

8. The dynamic allocation and management method for low-altitude airspace according to claim 4, characterized in that, In step S6, during the solution process using the improved genetic algorithm, an initial population is randomly generated. The initial population satisfies that each flight mission has one and only one dynamic airspace unit in each time slice, and the dynamic airspace units of each flight mission in adjacent time slices are adjacent. The comprehensive objective function is used as the fitness function, and a roulette wheel selection or tournament selection mechanism is adopted to select high-quality individuals from the current population as parent individuals to enter the mating pool based on individual fitness values. Design the following crossover operation to maintain track continuity: A. Randomly select two parent individuals from the mating pool. and From the flight mission set Randomly select a flight mission As an intersection object; B. Randomly determine a continuous time segment. , This range covers flight missions Part or all of the flight time; C. Parental individuals Mid-flight mission In time segment All allocation schemes within The entire entity is replaced with the parent individual. Mid-flight mission The allocation scheme within the same time segment; simultaneously, the parent individuals Mid-flight mission exist The internal allocation scheme was replaced with The original plan remains unchanged; the allocation of remaining flight missions and remaining time slices remains the same. D. Verify the feasibility of the offspring individuals generated by the crossover operation, and check at the exchange boundary, i.e. and Between and and In between, flight missions Does the trajectory still satisfy the spatial adjacency constraint? If not, make local adjustments to the adjacent dynamic spatial units at the boundary along the trajectory direction to restore the continuity of the trajectory. The adjustment method is: in the set of adjacent dynamic spatial units at the boundary time slice, select the dynamic spatial unit that minimizes the decrease in the comprehensive objective function as the transition unit. Design the following mutation operation that preserves neighborhood relationships: E. Select any one of the offspring individuals generated by the crossover operation. ,exist Randomly select a flight mission and a time slice , ; F. Obtaining Flight Missions In time slice Allocated airspace units Right now and time slices and Assigned dynamic spatial units and ; G. Determine the set of candidate variant spatial units, including the candidate variant spatial units in this set. Simultaneously satisfying the following three conditions: the candidate mutated spatial unit in the time slice Not occupied by other flight missions; from Departure, in a time slice Reachable ,Right now ;from Departure, in a time slice Reachable ,Right now ; H. If the candidate mutation airspace unit set is empty, then abandon this mutation and reselect the flight mission and time slice; if it is not empty, then randomly select a candidate mutation airspace unit from the candidate mutation airspace unit set. Flight mission In time slice The allocation from Change to , that is to say , ; I. Perform the above mutation operation on each offspring individual with a preset mutation probability.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the dynamic allocation and management method for low-altitude airspace as described in any one of claims 1 to 8.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the dynamic allocation and management method for low-altitude airspace as described in any one of claims 1 to 8.

Images