An intelligent airspace resource allocation method for low-altitude safety

Abstract

This invention discloses an intelligent airspace resource allocation method for low-altitude safety, relating to the field of airspace management technology. The method includes: collecting flight target status information and airspace environment status information in low-altitude airspace, and processing this information with unified time stamps and continuous trajectory association to form a basic airspace perception dataset; based on this dataset, identifying flight target behavior characteristics within continuous time intervals and determining their tendencies to form target intent determination information; processing the target intent determination information with spatial location correspondences to identify the risk level of each spatial location in the low-altitude region, and performing temporal mapping within continuous time intervals to obtain dynamic risk potential field information. This invention unifies the temporal changes in risk in the low-altitude region, airspace state switching, and flight response relationships into a single decision framework, creating a sequential constraint relationship between airspace allocation strategies and flight path strategies.

Description

Technical Field

[0001] This invention relates to the field of airspace management technology, and in particular to an intelligent airspace resource allocation method for low-altitude safety. Background Technology

[0002] In recent years, with the rapid development of the low-altitude economy, the application scenarios of drones in logistics, inspection, emergency rescue, and urban air traffic have been increasing, leading to a significant increase in the number of flying targets in low-altitude airspace. To improve the safety of low-altitude operations and the efficiency of airspace utilization, related technologies are gradually evolving from traditional static airspace planning to intelligent airspace management. Current research mainly focuses on multi-source perception fusion, flight trajectory prediction, drone intent recognition, and airspace collaborative scheduling. It utilizes integrated communication and sensing networks to achieve real-time acquisition of flight target status and environmental information. Furthermore, it combines multi-agent game models and swarm optimization algorithms to jointly decide on low-altitude flight path planning and airspace resource allocation, thereby supporting dynamic airspace management and collaborative drone operations in complex low-altitude environments.

[0003] However, in scenarios with dense low-altitude flight targets, existing technologies typically focus on single-dimensional path planning or airspace control strategies during airspace resource allocation. This makes it difficult to fully integrate multi-source information such as flight target behavior characteristics, intent evolution, and spatial risk distribution for collaborative decision-making. Consequently, there is a lack of an effective dynamic game-theoretic coordination mechanism between airspace allocation strategies and flight path strategies, affecting the overall adaptability and operational stability of airspace resource allocation in complex low-altitude environments. Therefore, how to achieve synergistic optimization of flight path strategies and airspace allocation strategies under dynamic risk environments has become a key technical problem that urgently needs to be solved in the field of intelligent low-altitude airspace management. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an intelligent airspace resource allocation method for low-altitude safety, which solves the problem of intelligent airspace resource allocation where flight path strategies and airspace allocation strategies are difficult to coordinate and optimize under complex low-altitude dynamic risk environments.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: This invention provides an intelligent airspace resource allocation method for low-altitude safety, comprising: collecting flight target status information and airspace environment status information in low-altitude airspace, and performing unified time labeling and continuous trajectory association to form an airspace perception basic dataset; based on the airspace perception basic dataset, identifying flight target behavior characteristics within a continuous time interval and performing tendency determination to form target intent determination information; performing spatial location correspondence processing on the target intent determination information, identifying the risk level of each spatial location in the low-altitude area, and performing temporal mapping within a continuous time interval to obtain dynamic risk potential field information; and based on the dynamic risk potential field information... A right-of-way migration matrix for reversible takeover of airspace is established, and a Stackelberg game model is constructed under constraints of blockade occupation, detour transition, and resumption of passage to determine airspace allocation strategies and flight path strategies, forming airspace game strategy information. The airspace game strategy information is decoupled in time and space, conflict closed-loop clusters are identified, and blockade pruning, detour insertion, and reset backfilling are performed to form an airspace resource allocation scheme. Based on the airspace resource allocation scheme, the flight scheduling and control of UAVs is driven. Real-time flight status information is continuously monitored through a 5G-A integrated sensing base station, and the airspace resource allocation scheme is dynamically adjusted based on feedback to form a dynamic airspace resource scheduling report.

[0007] As a preferred embodiment of the intelligent airspace resource allocation method for low-altitude safety described in this invention, the steps for forming the airspace perception basic dataset are as follows: The collected flight target status information and airspace environment status information are time-stamped and calibrated with a unified time axis to obtain a unified airspace observation sequence; By performing continuous trajectory connection and trajectory continuity correlation discrimination on the flight target state information in a unified airspace observation sequence, a continuous flight trajectory dataset is obtained; By spatially mapping the flight trajectory data and airspace environmental status information in the continuous flight trajectory dataset and fusing multi-source observation data, a basic dataset for airspace perception is generated.

[0008] As a preferred embodiment of the intelligent airspace resource allocation method for low-altitude safety described in this invention, the steps for identifying the behavioral characteristics of flying targets within a continuous time interval are as follows: The airspace perception basic dataset is sliced ​​into time intervals according to a preset continuous time window, and the flight trajectory data, neighboring flight target interaction data and airspace boundary constraint data are extracted from the corresponding time interval with the flight target as the center to obtain the target center time window data block; Recursive interaction relationships are constructed for the target center time window data blocks, and a time-expanded interaction relationship structure is established in chronological order to obtain a recursive interaction relationship sequence. By performing continuous path encoding and behavioral feature extraction on the recursive interaction sequence, the behavioral features of the flight target are obtained.

[0009] As a preferred embodiment of the intelligent airspace resource allocation method for low-altitude safety described in this invention, the steps for forming target intent determination information are as follows: The behavioral intensity of flight target behavior characteristics is assessed over a continuous time interval to determine behavioral tendency indicators and obtain flight behavior tendency information. Intent category mapping and consistency judgment are performed on flight behavior tendency information to form target intent determination information.

[0010] As a preferred embodiment of the intelligent airspace resource allocation method for low-altitude safety described in this invention, the steps for obtaining dynamic risk potential field information are as follows: The target intent determination information is organized according to spatial location, and the target intent determination information is associated with the corresponding spatial location to obtain a spatial location intent distribution dataset; Spatial clustering relationships are identified in the target intent determination information in the spatial location intent distribution dataset, and the risk level of the corresponding spatial location is marked to obtain a spatial risk label dataset; The risk levels in the spatial risk identification dataset are mapped and spatially correlated over a continuous time interval to generate dynamic risk potential field information.

[0011] As a preferred embodiment of the intelligent airspace resource allocation method for low-altitude safety described in this invention, the steps for constructing the Stackelberg game model are as follows: Perform risk transmission closure identification and risk return boundary locking on dynamic risk potential field information, divide it into reversible takeover airspace, and write the corresponding blockade occupation status, detour transition status and resumption release status into the risk status parameter set; Based on the risk status parameter set, the allowed switching order of adjacent reversible takeover airspaces in the closed-off, transit, and restored states is written into the correlation between each reversible takeover airspace to construct an airspace passage right migration matrix. The prevention and control party is defined as the initiator of the right-of-way migration action, and the drone operator is defined as the path connection response entity. A master-slave strategy space is established according to the airspace right-of-way migration matrix to obtain the strategy space dataset. A payoff function is constructed for the right-of-way migration action and path connection action in the policy space dataset. The payoff function is used to constrain the sequential switching relationship between the blockade occupation state, the detour transition state and the restoration release state, and a Stackelberg game model is established.

[0012] As a preferred embodiment of the intelligent airspace resource allocation method for low-altitude safety described in this invention, the steps for forming airspace game strategy information are as follows: Based on the Stackelberg game model, the payoff of right-of-way migration actions in the policy space dataset is evaluated according to the allowed switching order. Combinations of right-of-way migration actions that satisfy the sequential switching relationship are selected to determine the airspace partitioning strategy. The airspace partitioning strategy is input into the Stackelberg game model to limit the path connection response range of the drone operator, and the optimal path connection response action corresponding to the airspace partitioning strategy is solved according to the payoff function to determine the flight path strategy. The airspace division strategy and flight path strategy are combined and matched according to the reversible takeover airspace and corresponding time location, and then filtered according to the preset equilibrium conditions to form airspace game strategy information.

[0013] As a preferred embodiment of the intelligent airspace resource allocation method for low-altitude safety described in this invention, the steps for forming the airspace resource allocation scheme are as follows: The airspace division strategy in the airspace game strategy information is broken down into airspace occupation entries of each reversible takeover airspace at the corresponding time position, and the flight path strategy is broken down into path occupation entries of the flight target at the corresponding spatial position and the corresponding time position, forming a spatiotemporal occupation list. The spatial and temporal occupancy list is used to link and organize the mutual influence between airspace occupancy items and path occupancy items caused by the closed occupancy status, the transit status, and the resumption of passage status, forming a constraint write-back diagram. The mutual restraint relationships in the restraint and back-rewrite diagram are closed-loop merged, conflicting closed-loop clusters are identified, and blockade pruning, bypass insertion, and reset backfilling are performed to form an airspace resource allocation scheme.

[0014] As a preferred embodiment of the intelligent airspace resource allocation method for low-altitude safety described in this invention, the step of continuously monitoring real-time flight status information through a 5G-A integrated sensing base station includes the following steps: The airspace resource allocation scheme is scheduled and parsed and converted into UAV flight scheduling and control commands to drive the UAV to execute flight missions and obtain actual flight execution status data. By using a 5G-A integrated sensing base station, the actual flight execution status data is continuously sensed and collected to form real-time flight status information.

[0015] As a preferred embodiment of the intelligent airspace resource allocation method for low-altitude safety described in this invention, the step of generating a dynamic airspace resource scheduling report includes the following steps: The spatiotemporal deviation between real-time flight status information and flight path strategy in airspace resource allocation scheme is calculated to obtain airspace scheduling deviation index. Based on the airspace scheduling deviation index, the airspace resource allocation scheme is dynamically adjusted and the scheduling process information is recorded to form a dynamic airspace resource scheduling report.

[0016] The beneficial effects of this invention are as follows: A right-of-way migration matrix for reversible takeover airspace is established based on dynamic risk potential field information. A Stackelberg game model is constructed under constraints of blockade occupation, detour transition, and resumption of passage. This unifies the temporal changes in risk in low-altitude areas, airspace state switching, and flight response relationships into a single decision-making framework. This ensures that airspace allocation strategies and flight path strategies form a sequential constraint relationship, promoting orderly connection of right-of-way migration between different reversible takeover airspaces. Simultaneously, it clarifies the action and response boundaries of the control party and the drone operator within a continuous time interval, providing a continuous, clear, and executable strategic basis for decoupling spatiotemporal occupation, identifying conflict closed-loop clusters, and forming airspace resource allocation schemes. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart for an intelligent airspace resource allocation method oriented towards low-altitude safety.

[0019] Figure 2 A flowchart for generating spatial game strategy information.

[0020] Figure 3 A flowchart for generating dynamic risk potential field information.

[0021] Figure 4 A flowchart for generating the basic dataset for spatial awareness. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4 This is one embodiment of the present invention, which provides a smart airspace resource allocation method for low-altitude safety, comprising the following steps: S1. Collect flight target status information and airspace environment status information in low-altitude airspace, and organize them with unified time labels and continuous trajectory association to form a basic dataset for airspace perception.

[0026] S1.1: The collected flight target status information and airspace environment status information are processed with time stamps and a unified time axis is calibrated to obtain a unified airspace observation sequence; Specifically, flight target status information and airspace environment status information in the low-altitude area are collected by 5G-A integrated sensing base station, micro-radar and visual observation equipment, respectively. The collected flight target status information and airspace environment status information are then labeled with their respective time stamps and flight target identifiers. The time stamps in the flight target status information and the airspace environment status information are aligned with a unified time axis to form a one-to-one correspondence between the flight target status information and the airspace environment status information at the unified time stamp position. The aligned flight target status information and airspace environment status information are then continuously combined and arranged according to the unified time stamp order to obtain a unified airspace observation sequence.

[0027] It should be noted that the flight target status information includes the flight target's identification, real-time spatial position coordinates, flight speed, heading angle, and signal characteristics sensed by its 5G-A integrated sensing base station; Airspace environmental status information includes the distribution of equipment within and around the substation and the corresponding spatial location value (e.g., the core area of ​​the substation has high value, while ordinary areas have low value), pre-defined geographical boundaries and airspace functional zones (e.g., core equipment area, no-fly zone, and work area), the current airspace occupancy status (e.g., the number and distribution of drones), the substation's operating status and electromagnetic environment parameters, and meteorological environment data.

[0028] S1.2: Perform continuous trajectory connection and trajectory continuity correlation judgment on the flight target status information in the unified airspace observation sequence to obtain a continuous flight trajectory dataset; Specifically, the flight target status information in the unified airspace observation sequence is arranged sequentially according to a unified time identifier, and then collected into a continuous record sequence for the corresponding flight target according to the flight target identifier. In the continuous record sequence, the spatial position, flight speed, heading angle, and acquisition time interval under adjacent unified time identifiers are correlated to determine whether the subsequent spatial position can be reached from the previous spatial position within the corresponding time interval according to the current flight speed and heading, and whether the direction of the line connecting the heading changes, speed changes, and spatial positions maintains the same motion trend. When the status information of adjacent flight targets can be mutually verified in terms of time sequence, spatial reachability, speed continuity, and heading consistency, the status information of adjacent flight targets is connected into the same continuous trajectory. When there is a short-term missing record but the motion trend before and after the missing record can still be maintained, a trajectory missing mark is written and the original continuous trajectory relationship is maintained. When the status information of adjacent flight targets cannot be continuously verified in time, space, speed, or heading, the corresponding position is marked as a trajectory breakpoint, and a continuous trajectory record is re-established from the breakpoint. The continuous trajectory records corresponding to each flight target are collected to obtain a continuous flight trajectory dataset.

[0029] S1.3: Spatial location correspondence and multi-source observation data fusion are performed on the flight trajectory data and airspace environment status information in the continuous flight trajectory dataset to generate the basic dataset for airspace perception.

[0030] Specifically, the flight trajectory data in the continuous flight trajectory dataset is arranged in a unified time identifier order, and the spatial position records under the corresponding unified time identifier are extracted from the flight trajectory data. The spatial position records in the flight trajectory data are matched one by one with the spatial position ranges under the corresponding unified time identifiers in the airspace environment status information, so that the spatial position records in the flight trajectory data and the airspace environment status information form a corresponding record relationship under the same spatial position range and unified time identifier conditions. The flight trajectory data with corresponding record relationships and the airspace environment status information are combined and recorded in a unified time identifier order to generate the airspace perception basic dataset.

[0031] S2. Based on the basic dataset of airspace perception, identify the behavioral characteristics of flying targets within a continuous time interval, and determine the tendency of the flying target behavioral characteristics to form target intent determination information.

[0032] S2.1: The airspace perception basic dataset is sliced ​​into time intervals according to a preset continuous time window, and the flight trajectory data, the interaction data of the neighboring flight targets and the airspace boundary constraint data are extracted in the corresponding time interval with the flight target as the center to obtain the target center time window data block; Specifically, based on a preset continuous time window, the airspace perception basic dataset is divided into time intervals, so that combined records within the same continuous time window form a corresponding time interval record set. In the corresponding time interval record set, the flight trajectory data of the corresponding flight target is located according to the flight target identifier, so that the corresponding flight target forms a flight trajectory record sequence centered on the flight target in the corresponding time interval record set. Within the spatial location range corresponding to the flight trajectory record sequence, the flight trajectory data in the airspace perception basic dataset is located to form neighboring flight target interaction data. Within the spatial location range corresponding to the flight trajectory record sequence, the airspace environment state information in the airspace perception basic dataset is located to form airspace boundary constraint data. The flight trajectory data, neighboring flight target interaction data, and airspace boundary constraint data are combined and organized according to the corresponding time interval record set to obtain the target center time window data block.

[0033] It should be noted that a continuous time window is a range of continuous time intervals formed by selecting a fixed number of adjacent time markers in the order of the unified time markers, based on the adjacent arrangement relationship between the unified time markers.

[0034] S2.2: Construct recursive interaction relationships for the target center time window data blocks, and establish a time-expanded interaction relationship structure in chronological order to obtain a recursive interaction relationship sequence; Specifically, the target center time window data blocks are arranged according to a unified time identifier order, and an interactive correspondence between flight targets is established based on the flight trajectory data and the interaction data of neighboring flight targets in the target center time window data blocks, so that the flight trajectory data and the interaction data of neighboring flight targets form an interaction relationship record at the corresponding time position; the interaction relationship records at adjacent time positions are continuously associated and organized according to the unified time identifier order, so that the interaction relationship records form a continuously unfolded interaction relationship structure in time sequence; the interaction relationship structure is continuously arranged according to the unified time identifier order to obtain a recursive interaction relationship sequence.

[0035] S2.3: Perform continuous path encoding and behavioral feature extraction on the recursive interaction sequence to obtain the flight target behavior features.

[0036] Specifically, the interaction records between flight targets in the recursive interaction sequence are connected and organized into continuous paths according to a unified time identifier order, so that the interaction records of the same flight target in the recursive interaction sequence form continuous path records according to a unified time identifier order. In the continuous path records, the interaction records corresponding to adjacent time positions are organized segment by segment according to a unified time identifier order, so that the interaction records at adjacent time positions form interaction change records. The interaction change records are then arranged continuously according to a unified time identifier order to obtain the flight target behavior characteristics.

[0037] S2.4: Evaluate the behavioral intensity of the flight target's behavior characteristics over a continuous time interval, determine the behavioral tendency index, and obtain flight behavior tendency information; Specifically, within a continuous time interval, the flight target behavior characteristics corresponding to each unified time marker position are processed. If the number of times the flight target behavior characteristics appear within the time interval reaches a preset threshold, and the duration of each appearance exceeds a preset duration threshold, then the flight target behavior characteristics are marked as strong behavior and assigned a high behavior tendency index. If the number of times the flight target behavior characteristics appear within the time interval does not reach the preset threshold, and the duration of each appearance is lower than the preset duration threshold, then the flight target behavior characteristics are marked as weak behavior and assigned a low behavior tendency index, thus forming flight behavior tendency information.

[0038] It should be noted that the duration threshold is set based on the duration distribution of flight target behavior characteristics in actual application scenarios and the importance of the behavior characteristics. The specific setting steps include: determining the duration distribution of flight target behavior based on flight target behavior data, selecting a time interval that can effectively identify continuous behavior, and selecting the median value within the time interval as the preset duration threshold; the exemplary value range is 5 seconds to 30 seconds. When the duration is less than 5 seconds, it is impossible to distinguish between normal behavior and occasional short-term behavior; when the duration is greater than 30 seconds, it leads to over-identification of normal behavior and ignoring abnormal behavior in a short period of time. The frequency threshold is set based on the common occurrence frequency of flight target behavior characteristics within a certain time interval. The specific setting steps include: through statistical analysis of flight target behavior data, selecting a frequency that can effectively distinguish between frequent and occasional behavior as the preset frequency threshold; the exemplary value range is 3 to 10 times. When the frequency of flight target behavior is less than 3 times, the continuity and regularity of the behavior cannot be effectively identified; when the frequency is more than 10 times, frequent normal behavior will be misjudged as abnormal behavior.

[0039] S2.5: Map the flight behavior tendency information to the intention category and determine consistency to form target intention determination information.

[0040] Specifically, flight behavior tendency information is organized according to a unified time identifier sequence, and risk intention categories are determined by combining heading changes, stationary status, spatial position, relative relationship with adjacent flight targets, and airspace boundary constraints. Restricted areas refer to the boundaries of no-fly zones, core equipment areas, and operational areas requiring control as pre-defined in the airspace environmental status information. Frequent heading changes refer to the same flight target deviating from the continuous flight trajectory direction multiple times within a continuous time window. Long-term loitering refers to the same flight target remaining in the same spatial location range or lingering in a small area under multiple unified time identifiers. Approaching high-value spatial locations refers to the flight target entering the core equipment area, key facility area, or safety buffer zone. Abnormal interaction refers to the flight target continuously shortening the distance with adjacent flight targets, crossing and approaching, tailing, or gathering around them. The above behaviors are mapped to high-risk intention categories. Short-term deviations, temporary detours, and passage near the boundary caused by airspace boundary constraints or path adjustments are mapped to transitional risk intention categories. Behaviors where spatial position, heading, and speed are consistent with the continuous flight trajectory direction are mapped to low-risk passage intention categories. Consistent risk intent categories for the same flight target within adjacent consecutive time intervals are determined. If the current risk intent category is the same as that of the adjacent time interval, or changes gradually in the order of adjacent levels from low-risk passage intent category to transitional risk intent category to high-risk intent category, the risk intent category of the current time interval is retained. If the current time interval experiences a jump across levels, and the previous and subsequent time intervals do not show the same or adjacent level changes, the current time interval is marked as an isolated abrupt change, and is corrected by combining the flight behavior tendency information of the preceding and following time intervals to form target intent determination information.

[0041] S3. Organize the spatial location correspondence of the target intent determination information, identify the risk level of each spatial location in the low-altitude area, and perform time-series mapping in a continuous time interval to obtain dynamic risk potential field information.

[0042] S3.1: Organize the target intent determination information according to spatial location, and associate the target intent determination information with the corresponding spatial location to obtain the spatial location intent distribution dataset; Specifically, the target intent determination information is arranged in a unified time identifier order, and the spatial position of the corresponding flight target at the unified time identifier position is determined in the target intent determination information based on the flight trajectory data, so that the target intent determination information and the spatial position in the flight trajectory data are corresponding. The target intent determination information and the corresponding spatial position are linked and recorded one by one in a unified time identifier order, so that each target intent determination information corresponds to a spatial position. The target intent determination information after spatial position correspondence is sorted and collected according to spatial position to obtain a spatial position intent distribution dataset.

[0043] S3.2: Identify spatial clustering relationships in the target intent determination information in the spatial location intent distribution dataset, and label the risk level of the corresponding spatial location to obtain a spatial risk label dataset; Specifically, based on the spatial location intent distribution dataset, the concentration of target intent judgment information within adjacent spatial locations is examined. This is combined with the flight target type, the distance between the corresponding spatial location and no-fly zones, core equipment areas, key facility areas, flight speed, altitude changes, and airspace boundary constraints to determine the risk level corresponding to spatial clustering relationships. When high-risk intent categories appear in clusters, and the flight targets are close to no-fly zones and core equipment areas, a high-risk level label is assigned. When transitional-risk intent categories appear in clusters, and the corresponding spatial locations are located at the boundaries of the work area, geographical boundaries, and adjacent airspace functional zones, a medium-risk level label is assigned. When low-risk passage intent categories appear in clusters, and the flight speed, altitude, heading angle, and flight target label all match the airspace functional zone, geographical boundary, and airspace occupancy status, a low-risk level label is assigned. This process yields a spatial risk label dataset.

[0044] S3.3: The risk levels in the spatial risk identification dataset are mapped and spatially correlated over a continuous time interval to generate dynamic risk potential field information.

[0045] Specifically, the risk level identifiers in the spatial risk identifier dataset are arranged in a continuous time interval according to the order of a unified time identifier. Based on the proximity of spatial locations, the risk level identifiers of each spatial location under adjacent unified time identifiers are connected and compared to form a sequence of risk level changes over time for the same spatial location and a distribution relationship of risk level for different spatial locations under the same unified time identifier. The risk level change sequence and the risk level distribution relationship are combined and recorded to generate dynamic risk potential field information.

[0046] S4. Based on dynamic risk potential field information, establish a right-of-way migration matrix for reversible takeover airspace, and construct a Stackelberg game model under the constraints of blockade occupation, detour transition, and resumption of passage to determine airspace division strategy and flight path strategy, thereby forming airspace game strategy information.

[0047] S4.1: Perform risk transmission closure identification and risk return boundary locking on dynamic risk potential field information, divide it into reversible takeover airspace, and write the corresponding blockade occupation status, detour transition status and resumption release status into the risk status parameter set; The risk level change sequence of the same spatial location in the dynamic risk potential field information, and the risk level distribution relationship of adjacent spatial locations under the same unified time identifier, are connected in sequence according to time and spatial proximity to form a risk transmission chain. When the risk level change direction of adjacent spatial locations in the risk transmission chain is continuous and consistent, and the risk location that meets the closure identification condition extends continuously outward from the starting spatial location, and then points back to the starting spatial location or the vicinity through the adjacent spatial location, the risk transmission closure is determined. In the risk transmission closure, the location where the risk level change direction changes from outward extension to return convergence is locked, and the adjacency boundary between the corresponding location and the previous spatial location is determined as the risk return boundary. Using the risk return boundary as a separator, the area within the boundary where the risk level reaches the takeover threshold and the spatial locations are continuously adjacent is merged into the reversible takeover airspace. Based on the continuous risk level status, passage connectivity status, and restriction release status of the reversible takeover airspace in the continuous time interval, the risk status is written into the blockade occupation status, the transit transition status, and the resumption of passage status, respectively, to form a risk status parameter set.

[0048] It should be noted that the closure identification threshold is a criterion for judging whether a risk transmission chain possesses the characteristic of "continuous extension and retracement closure." It should not be understood as a single numerical value, but rather should be jointly determined by the lower limit of risk level, the lower limit of duration, the lower limit of spatial adjacency quantity, and the upper limit of retracement distance. The risk level values ​​corresponding to low-risk, medium-risk, and high-risk level identifiers are extracted from the spatial risk identifier dataset. The adjacency range of adjacent spatial locations is determined based on the spatial division accuracy and airspace functional zoning in the airspace environmental status information. The duration requirement is determined by combining this with a continuous time window, specifying that "risk levels reaching medium risk or above..." The following criteria are set as the closure identification thresholds: "continuous duration of at least one continuous time window, at least two consecutive spatial locations extending along adjacent spatial locations, and the retracement position falling within the range of the starting spatial location or the adjacent spatial location of the starting spatial location". For example, when the risk level is low risk 1, medium risk 2, or high risk 3, the following criteria are set as the closure identification thresholds: risk level not lower than 2, continuous duration of at least 5 to 30 seconds, number of consecutive adjacent spatial locations not less than 2, and retracement distance not exceeding one spatial adjacent unit. This avoids misjudging occasional single-point risk fluctuations as risk transmission closure. The takeover threshold is set based on the risk level value, spatial location value, continuity of risk between adjacent spatial locations, and duration within a continuous time interval from the dynamic risk potential field information. The setting steps include: extracting the risk level values ​​corresponding to low-risk, medium-risk, and high-risk level identifiers from the spatial risk identifier dataset; determining the spatial location value by combining the core equipment area, no-fly zone, work area, and ordinary area from the airspace environmental status information; statistically analyzing the duration for which adjacent spatial locations continuously reach the same risk level within a continuous time interval; and using "risk level reaching medium risk or above, spatial location having the necessity for passage control, at least two consecutive adjacent spatial locations, and duration at least one continuous time window" as the basic threshold for reversible takeover airspace. For example, when the risk level value is set as low risk 1, medium risk 2, and high risk 3, the selected duration threshold within the range of risk level not lower than 2, spatial location value not lower than the preset medium value level, number of consecutive adjacent spatial locations not less than 2, and duration not less than 5 to 30 seconds can be used as the division condition. After the risk transmission chain completes its continuous outward extension, the direction of risk level change begins to converge and reverts to the spatial position corresponding to the initial spatial position or the area adjacent to the initial spatial position. Return convergence refers to the situation where, after the risk transmission chain extends outward, the distance between subsequent spatial locations and the initial spatial location continuously decreases, and the risk level of the corresponding spatial location still reaches the closure identification threshold, indicating that the direction of risk change has shifted from moving away from the initial spatial location to moving closer to the initial spatial location.

[0049] S4.2: Based on the risk status parameter set, the allowed switching order of adjacent reversible takeover airspaces in the closed-off, transit, and restored states is written into the correlation between each reversible takeover airspace to construct an airspace passage right migration matrix. The reversible takeover airspaces in the risk status parameter set are matched one by one according to spatial adjacency and time sequence in continuous time intervals. The time positions of the blockade occupation status, the transit status, and the resumption of passage status are marked between each pair of adjacent reversible takeover airspaces. The connection order of the blockade occupation status, the transit status, and the resumption of passage status between adjacent reversible takeover airspaces is checked according to the time position. The blockade occupation status in the previous time position is matched with the transit status in the next time position and written into the association relationship. The transit status in the previous time position is matched with the resumption of passage status in the next time position and written into the association relationship. The state order that can maintain continuous connection is recorded as the allowed switching order. By collecting each pair of adjacent reversible takeover airspace and the corresponding allowed switching order, an airspace right-of-way migration matrix is ​​constructed.

[0050] S4.3: Define the prevention and control party as the initiator of the right-of-way migration action, define the drone operator as the path connection response entity, establish a master-slave strategy space according to the airspace right-of-way migration matrix, and obtain the strategy space dataset; Specifically, based on the adjacent reversible takeover airspaces and the allowed switching order in the airspace right-of-way migration matrix, the corresponding entity that issues the right-of-way migration announcement for the closed-off occupation state, the transit state, and the restored release state is defined as the prevention and control party, and the corresponding entity that completes the flight path connection in the corresponding reversible takeover airspace and the corresponding time position according to the allowed switching order is defined as the drone operator. The control party's right-of-way transfer actions between adjacent reversible takeover airspaces are taken as the primary strategy, and the drone operator's path connection response actions under the right-of-way transfer actions are taken as the secondary strategy. Each adjacent reversible takeover airspace, corresponding time and location, allowed switching order, right-of-way transfer actions and path connection response actions are recorded item by item to establish a master-slave strategy space and obtain the strategy space dataset.

[0051] S4.4: Construct payoff functions for right-of-way migration actions and path connection actions in the policy space dataset, and use the payoff functions to constrain the sequential switching relationship between the blockade occupation state, the detour transition state, and the restoration release state, and establish a Stackelberg game model.

[0052] Specifically, when constructing a Stackelberg game model for right-of-way migration actions and path connection response actions in the strategy space dataset, the control party is considered the leader, and the drone operator is considered the follower. The right-of-way migration actions chosen by the control party in each reversible takeover airspace and corresponding time location—such as blockade occupation, detour transition, and resumption of passage—are considered as the leader's strategy variables. The path connection response actions chosen by the drone operator under the leader's strategy constraints—such as avoidance, detour, waiting, and resetting passage—are considered as the follower's strategy variables. The control party's strategy is constructed using the risk level, spatial location value, state consistency identifier, and allowed switching order in the dynamic risk potential field information. The project establishes a safety benefit function; constructs a cost function for drone operators based on flight path length, total flight time, risk airspace occupancy, and path continuity constraints; and assigns sequential correspondences to the first blocked occupancy state, the second detour transition state, and the subsequent resumption of passage state according to the master-slave correspondence recorded in the benefit function. It removes right-of-way transfer actions and path connection actions that do not conform to the sequence from the master-slave correspondence, while retaining those that do conform to the sequence. The project uses the benefit function to constrain the sequential switching relationship between the blocked occupancy state, the detour transition state, and the resumption of passage state, and establishes a Stackelberg game model.

[0053] It should be noted that the Stackelberg game model consists of a subject layer, a state constraint layer, a strategy variable layer, an objective function layer, an optimal response layer, and an equilibrium output layer. The subject layer treats the control party as the leader and the drone operator as the follower. The state constraint layer consists of reversible takeover of airspace, airspace right-of-way migration matrix, blockade and occupation state, detour transition state, and restoration of clearance state. The strategy variable layer uses the control party's right-of-way migration action as the leader's strategy variable and the drone operator's path connection response action as the follower's strategy variable. The objective function layer constructs the control party's safety benefit function and the drone operator's cost function, respectively. In the Stackelberg game model, the defense party corresponds to the safety benefit function, and the drone operator corresponds to the cost function. In the stage of selecting the combination of airspace allocation strategy and flight path strategy, the safety benefit function and cost function are merged into a collective benefit function.

[0054] The expression for the safety benefit function is: ; The expression for the cost function is: ; The payoff function used when combining and selecting airspace allocation strategies and flight path strategies is: ; in, This represents the value of the safety benefit function; Indicates the number of reversible takeover airspace; Indicates the reversible takeover airspace number; Indicates the first The risk level value corresponding to each reversible takeover airspace is derived from the risk indicator of the corresponding spatial location in the continuous time interval in the dynamic risk potential field information (for example, the high risk indicator is assigned a value of 3, the medium risk indicator is 2, and the low risk indicator is 1). Indicates the first The spatial location value corresponding to a reversible takeover airspace; Indicates the first A consistent identifier for the status of a reversible takeover airspace; Represents the cost function value; , and Indicates the weighting coefficient; This represents the total length of the flight path corresponding to the flight path strategy. Indicates the maximum flight path length; This indicates the total flight time corresponding to the flight path strategy; Indicates the maximum flight time; Indicates the flight path strategy for the first An occupancy marker for a reversible takeover airspace; This represents the value of the payoff function.

[0055] It should be noted that the state consistency identifier originates from the first... The reversible takeover airspace is recorded in the risk status parameter set as either under lockdown, transit, or reopening, and is related to the first... The corresponding records between the risk identifier of each reversible takeover airspace in the dynamic risk potential field information and the allowed switching order in the airspace right-of-way migration matrix; when the first The value is one when the status recorded for a reversible takeover airspace is consistent with the corresponding risk identifier and the allowed switching order; and the value is zero when it is inconsistent with the corresponding risk identifier and the allowed switching order. In the safety return function, it only plays a role of "including or not including": when Take one moment, the first The risk level and spatial location value corresponding to each reversible takeover airspace are jointly recorded in the safety benefit function; when When zero is taken, the first The corresponding terms of a reversible takeover space domain are not included in the safety benefit function; The occupancy identifier is derived from the path occupancy entries after the flight path strategy is broken down and the first... The corresponding records of each reversible takeover airspace at the corresponding spatial and temporal locations; when the flight path strategy passes through, enters, or stays in the first... When a reversible takeover airspace is encountered, the value is set to one, and the flight path strategy has not passed through, entered, or remained in the first reversible takeover airspace. When a reversible takeover space is encountered, the value is zero; In the cost function, with the first The risk levels corresponding to each reversible takeover airspace occur simultaneously; when Take one moment, the first The risk level corresponding to each reversible takeover airspace is recorded in the cost function; when When zero is taken, the first The risk level corresponding to a reversible takeover airspace is not included in the cost function; The maximum flight path length is derived from the geographical boundaries, airspace functional zones, and passable space range recorded in the airspace environmental status information, as well as the allowed switching order recorded in the airspace right-of-way migration matrix. The non-passable sections corresponding to the blockade and occupation status are excluded between the starting and target spatial positions, while the continuous passable sections corresponding to the transitional and reopened states are retained. The entire allowable passage after connecting the beginning and end of the continuous passable sections is used as the boundary record of the flight path length. The maximum flight duration is derived from the start time position, end time position, and allowed switching order of the corresponding continuous passable section within a continuous time interval; the inaccessible time position corresponding to the closed-off occupation state is excluded, the continuous passable time position corresponding to the transitional state and the restored passable state is retained, and the entire allowed passable time period from the start time position to the end time position is used as the boundary record of the flight duration. The weighting coefficients are determined based on the occurrence of flight trajectory data within each risk area block in the continuous flight trajectory dataset and the airspace functional zoning in the airspace environmental status information. High-importance risk area blocks are assigned the highest weighting coefficient, medium-importance risk area blocks are assigned the middle weighting coefficient, and low-importance risk area blocks are assigned the lowest weighting coefficient. The weighting coefficients for all risk area blocks are then adjusted proportionally to ensure a stable correspondence between the weighting coefficients of each risk area block, and to satisfy… .

[0056] S4.5: Based on the Stackelberg game model, evaluate the payoff of right-of-way migration actions in the policy space dataset according to the allowed switching order, select the combination of right-of-way migration actions that satisfy the sequential switching relationship, and determine the airspace partitioning strategy. Specifically, each right-of-way migration action in the strategy space dataset is matched item by item according to the adjacent reversible takeover airspace, the corresponding time position, and the allowed switching order in the airspace right-of-way migration matrix. In the Stackelberg game model, combined with the master-slave correspondence recorded in the payoff function, each right-of-way migration action is compared to see if the sequential switching relationship of the blockade occupation state comes first, the transit state in the middle, and the resumption of passage state comes last. Right-of-way migration action combinations that are consistent with the allowed switching order and can keep the adjacent reversible takeover airspace connected continuously in the continuous time interval are selected as candidate right-of-way migration action combinations. Right-of-way migration actions that do not meet the requirements of the blockade occupation state coming first, the transit state in the middle, and the resumption of passage state coming last are removed from the candidate right-of-way migration action combinations. The safety payoff function value corresponding to the candidate right-of-way migration action combination is compared with the safety payoff function, and the candidate right-of-way migration action combination with the highest safety payoff function value is retained as the airspace partitioning strategy.

[0057] S4.6: Input the airspace partitioning strategy into the Stackelberg game model to limit the path connection response range of the drone operator, and solve the optimal path connection response action corresponding to the airspace partitioning strategy based on the payoff function to determine the flight path strategy. Specifically, the airspace division strategy is incorporated into the Stackelberg game model, corresponding to each reversible takeover airspace, corresponding time location, blockade and occupation status, detour transition status, and resumption of passage status. The path connection actions in the reversible takeover airspace corresponding to the blockade and occupation status are excluded from the path connection response range of the drone operator, while the path connection actions in the reversible takeover airspace corresponding to the detour transition status and resumption of passage status are retained within the path connection response range of the drone operator. Each path connection action retained within the path connection response range is matched with the airspace division strategy item by item according to the reversible takeover airspace, corresponding time position, and allowed switching order. The cost function value corresponding to each path connection action is compared item by item in combination with the cost function. The path connection action with the smallest cost function value that simultaneously satisfies the following conditions is selected as the optimal path connection response action corresponding to the airspace division strategy, thus obtaining the flight path strategy.

[0058] S4.7: Combine and match airspace division strategies and flight path strategies according to reversible takeover airspace and corresponding time positions, and filter them according to preset equilibrium conditions to form airspace game strategy information.

[0059] Specifically, the reversible takeover airspace, corresponding time and location, blockade and occupation status, detour transition status, and resumption of passage status in the airspace division strategy are matched item by item with the corresponding reversible takeover airspace, corresponding time and location, and path connection response actions in the flight path strategy. Combinations are then made based on the same reversible takeover airspace and the same corresponding time and location, ensuring that the location corresponding to the blockade and occupation status in the airspace division strategy does not correspond to the path connection response action in the flight path strategy, and that the locations corresponding to the detour transition status and resumption of passage status in the airspace division strategy are connected sequentially with the path connection response actions in the flight path strategy. The combined airspace division strategies and flight path strategies are then filtered item by item against preset equilibrium conditions, retaining strategy combinations that simultaneously satisfy the following conditions: the control party's right-of-way transfer action remains in the first position; the drone operator's path connection response action remains in the second position; the blockade and occupation status is first; the detour transition status is in the middle; and the resumption of passage status is last, with each corresponding time and location continuously connected. This forms the airspace game strategy information.

[0060] It should be noted that the equilibrium condition refers to the judgment condition used to select the Stackelberg equilibrium result from the candidate strategy combination; each airspace allocation strategy is used as a candidate strategy for the control party, the executable flight path strategy under the strategy constraint is selected, and the cost function value of the UAV operator corresponding to each flight path strategy is calculated. The flight path strategy with the smallest cost function value is selected as the optimal path response; after combining each airspace allocation strategy and the corresponding optimal path response, the safety benefit function value of the control party is calculated. Under the premise that the blockade occupation state, the detour transition state and the resumption of release state are connected continuously in the allowed switching order, the strategy combination with the largest safety benefit function value is selected as the airspace game strategy information that satisfies the equilibrium condition.

[0061] S5. Decouple the spatial and temporal occupation of airspace game strategy information, identify conflict closed-loop clusters, and perform blockade pruning, bypass insertion and reset backfilling to form an airspace resource allocation scheme.

[0062] S5.1: The airspace division strategy in the airspace game strategy information is broken down into airspace occupation entries for each reversible takeover airspace at the corresponding time position, and the flight path strategy is broken down into path occupation entries for the flight target at the corresponding spatial position and the corresponding time position, forming a spatiotemporal occupation list; Specifically, the airspace division strategy in the airspace game strategy information is expanded item by item according to each reversible takeover airspace and its corresponding time position, and the closed-off occupation state, transit state, or resumption of passage state of each reversible takeover airspace at each corresponding time position is recorded as an airspace occupation item; the flight path strategy is expanded item by item according to the flight target, its corresponding spatial position, and its corresponding time position, and the path connection response action of the flight target at each corresponding spatial position and each corresponding time position is recorded as a path occupation item; the airspace occupation items and path occupation items are arranged in order of their corresponding time positions, and the reversible takeover airspace and its corresponding spatial position are aggregated at the same corresponding time position to form a spatiotemporal occupation list.

[0063] S5.2: The interrelationships between the airspace occupation items and the path occupation items in the space-time occupation list caused by the blockade occupation status, the detour transition status and the resumption of passage status are linked and organized to form a constraint write-back diagram; Specifically, the airspace occupancy entries corresponding to the closed-off occupancy status in the space-time occupancy list are written as restriction associations with the path occupancy entries that cannot pass under the same corresponding time position; the airspace occupancy entries corresponding to the transit status are written as transit associations with the path occupancy entries that need to change the passage position under the adjacent corresponding time positions; and the airspace occupancy entries corresponding to the restored passage status are written as passage associations with the path occupancy entries that re-enter the passage position under the corresponding time position. The restriction association, transition association, and release association are sequentially linked according to their corresponding time positions and the connection relationship between reversible takeover airspaces. The restriction, transition, and release relationship of the previous airspace occupation entry to the subsequent path occupation entry is written back between the corresponding entries to form a constraint write-back diagram.

[0064] S5.3: Close the loop and merge the mutual restraint relationships of the beginning and end of the restraint write-back diagram, identify conflicting closed loop clusters, and perform blockade pruning, bypass insertion and reset backfilling to form an airspace resource allocation scheme.

[0065] Specifically, based on the constraint write-back graph, when the subsequent association corresponds to the space occupancy entry or path occupancy entry corresponding to the previous association, the mutual constraint relationship of the beginning and end is gathered into the same closed loop, and the space occupancy entries and path occupancy entries in the same closed loop are merged into a conflict closed loop cluster. For airspace occupancy entries in the conflict loop cluster that were already connected to the release association at the previous corresponding time position, and were connected to the transition association at the next corresponding time position, and still maintained the blocked occupancy state, delete them from the conflict loop cluster. Only retain the airspace occupancy entries corresponding to the blocked occupancy state that still need to be associated with the restriction association. For path occupancy entries that cannot be continuously connected between adjacent corresponding time positions after trimming, write them sequentially into the reversible takeover airspace and corresponding time position where the airspace occupancy entry in the transition state is located, so that the path occupancy entries at the previous corresponding time position and the path occupancy entries at the next corresponding time position are continuously connected through the transition state. Change the inserted path occupancy entries back to their original passage positions at the reversible takeover airspace and corresponding time position corresponding to the restored release state, reconnect them with the path occupancy entries before and after the restored release state, and rearrange them according to the reversible takeover airspace and corresponding time position to form an airspace resource allocation scheme.

[0066] S6. Driving UAV flight scheduling and control based on airspace resource allocation scheme, continuously monitoring real-time flight status information through 5G-A integrated sensing base station, and dynamically adjusting the airspace resource allocation scheme to generate a dynamic airspace resource scheduling report.

[0067] S6.1: The airspace resource allocation scheme is scheduled and parsed and converted into UAV flight scheduling and control commands to drive the UAV to execute flight missions and obtain actual flight execution status data; Specifically, the spatial area, time window, sequential decision rules specified by the airspace allocation strategy and the spatial location, timestamp, and heading specified by the flight path strategy are extracted from the airspace resource allocation scheme and converted into UAV flight scheduling and control commands. The UAV flight scheduling and control commands are then sent to the designated UAVs through the UAV flight control platform, driving the UAVs to execute flight missions according to the UAV flight scheduling and control commands. During the execution of the flight mission, the UAV flight status data is recorded to obtain the actual flight execution status data.

[0068] S6.2: Through the 5G-A integrated sensing base station, the actual flight execution status data is continuously sensed and collected to form real-time flight status information.

[0069] Specifically, the 5G-A integrated sensing base station is used to perceive the actual flight status data of the UAV during the flight mission in real time. The 5G-A integrated sensing base station uses communication signals to continuously detect and measure the UAV, collect the UAV's real-time position, altitude, speed, heading and timestamp status data during flight, and use the 5G-A integrated sensing base station to transmit and aggregate the collected status data through the communication link to form real-time flight status information that includes the spatial position, motion status and time stamp of the flight target at a specific moment.

[0070] S6.3: Calculate the spatiotemporal deviation between real-time flight status information and flight path strategy in airspace resource allocation scheme to obtain airspace scheduling deviation index; Specifically, the actual spatial position of the UAV at a specific moment is extracted from the real-time flight status information and compared with the planned spatial position of the UAV at the corresponding moment in the flight path strategy to obtain the deviation distance between the actual spatial position and the planned spatial position in the spatial dimension. At the same time, the actual arrival time of the specific spatial position in the real-time flight status information is compared with the planned arrival time in the flight path strategy to obtain the deviation duration in the time dimension. Based on the deviation distance in the spatial dimension and the deviation duration in the time dimension, the deviation distance in the spatial dimension and the deviation duration in the time dimension are merged into a single quantitative indicator, namely the airspace scheduling deviation index, through normalized weighted summation.

[0071] S6.4: Based on the airspace scheduling deviation index, perform dynamic feedback adjustments to the airspace resource allocation scheme and record scheduling process information to generate an airspace resource dynamic scheduling report.

[0072] Specifically, the airspace resource allocation scheme currently in operation is dynamically adjusted based on the airspace scheduling deviation index. When the real-time flight status information is inconsistent with the planned spatial location, planned arrival time, planned speed, or planned heading in the flight path strategy, the airspace scheduling deviation index, real-time flight status information, deviation time and location, deviation spatial location, and the current airspace resource allocation scheme are written into the feedback adjustment record. Based on the feedback adjustment record, the real-time flight status information is written back to the airspace perception basic dataset, and the target intent determination information, dynamic risk potential field information, and risk status parameter set for the corresponding spatial location are updated. Based on the updated dynamic risk potential field information, the reversible takeover airspace, airspace partitioning strategy, and flight path strategy are redefined, and the airspace game strategy information is subjected to spatiotemporal occupancy decoupling, conflict closed-loop cluster identification, blockade pruning, bypass insertion, and reset backfilling to generate an updated airspace resource allocation scheme. During the scheduling cycle, the airspace scheduling deviation index, real-time flight status information, feedback adjustment record, airspace resource allocation scheme before and after adjustment, and actual execution status after adjustment are continuously recorded and summarized in chronological order to form an airspace resource dynamic scheduling report.

[0073] In summary, this invention establishes a right-of-way migration matrix for reversible takeover airspace based on dynamic risk potential field information, and constructs a Stackelberg game model under constraints of blockade occupation, detour transition, and resumption of passage. This unifies the temporal changes in risk in low-altitude areas, airspace state switching, and flight response relationships into a single decision-making framework, enabling airspace allocation strategies and flight path strategies to form a sequential constraint relationship. This promotes orderly connection of right-of-way migration between different reversible takeover airspaces, clarifies the action and response boundaries of the control party and the drone operator within a continuous time interval, and provides a continuous, clear, and executable strategic basis for decoupling spatiotemporal occupation, identifying conflict closed-loop clusters, and forming airspace resource allocation schemes.

[0074] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A low-altitude safety-oriented intelligent airspace resource allocation method, characterized in that, include: Collect flight target status information and airspace environment status information in low-altitude airspace, and organize them with unified time labels and associate them with continuous trajectories to form a basic dataset for airspace perception. Based on the basic dataset of airspace perception, the behavioral characteristics of flying targets within a continuous time interval are identified, and the tendency is determined to form target intent determination information; The spatial location correspondence of the target intent determination information is organized to identify the risk level of each spatial location in the low-altitude area, and time-series mapping is performed in a continuous time interval to obtain dynamic risk potential field information. Based on dynamic risk potential field information, a right-of-way migration matrix for reversible takeover airspace is established. Under the constraints of blockade occupation, detour transition, and resumption of passage, a Stackelberg game model is constructed to determine airspace allocation strategy and flight path strategy, thereby forming airspace game strategy information. The spatial and temporal occupancy information of airspace game strategy is decoupled, conflict closed-loop clusters are identified, and blockade pruning, bypass insertion and reset backfilling are performed to form an airspace resource allocation scheme. Driving UAV flight scheduling and control based on airspace resource allocation scheme, the system continuously monitors real-time flight status information through 5G-A integrated sensing base station, and dynamically adjusts the airspace resource allocation scheme to generate a dynamic airspace resource scheduling report. 2.The low-altitude safety oriented intelligent air space resource allocation method of claim 1, wherein, The steps for forming the basic dataset for spatial awareness are as follows: The collected flight target status information and airspace environment status information are time-stamped and calibrated with a unified time axis to obtain a unified airspace observation sequence; By performing continuous trajectory connection and trajectory continuity correlation discrimination on the flight target status information in a unified airspace observation sequence, a continuous flight trajectory dataset is obtained; By spatially mapping the flight trajectory data and airspace environmental status information in the continuous flight trajectory dataset and fusing multi-source observation data, a basic dataset for airspace perception is generated. 3.The low-altitude safety oriented intelligent air space resource allocation method of claim 2, wherein, The steps for identifying the behavioral characteristics of flying targets within a continuous time interval are as follows: The airspace perception basic dataset is sliced ​​into time intervals according to a preset continuous time window, and the flight trajectory data, neighboring flight target interaction data and airspace boundary constraint data are extracted from the corresponding time interval with the flight target as the center to obtain the target center time window data block; Recursive interaction relationships are constructed for the target center time window data blocks, and a time-expanded interaction relationship structure is established in chronological order to obtain a recursive interaction relationship sequence. By performing continuous path encoding and behavioral feature extraction on the recursive interaction sequence, the behavioral features of the flight target are obtained. 4.The low-altitude safety oriented intelligent air space resource allocation method of claim 3, wherein, The steps for forming the target intent determination information are as follows: The behavioral intensity of flight target behavior characteristics is assessed over a continuous time interval to determine behavioral tendency indicators and obtain flight behavior tendency information. Intent category mapping and consistency judgment are performed on flight behavior tendency information to form target intent determination information.

5. The intelligent airspace resource allocation method for low-altitude safety as described in claim 4, characterized in that, The steps to obtain the dynamic risk potential field information are as follows: The target intent determination information is organized according to spatial location, and the target intent determination information is associated with the corresponding spatial location to obtain a spatial location intent distribution dataset; Spatial clustering relationships are identified in the target intent determination information in the spatial location intent distribution dataset, and the risk level of the corresponding spatial location is marked to obtain a spatial risk label dataset; The risk levels in the spatial risk identification dataset are mapped and spatially correlated over a continuous time interval to generate dynamic risk potential field information.

6. The intelligent airspace resource allocation method for low-altitude safety as described in claim 1, characterized in that, The steps for constructing the Stackelberg game model are as follows: Perform risk transmission closure identification and risk return boundary locking on dynamic risk potential field information, divide it into reversible takeover airspace, and write the corresponding blockade occupation status, detour transition status and resumption release status into the risk status parameter set; Based on the risk status parameter set, the allowed switching order of adjacent reversible takeover airspaces in the closed-off, transit, and restored states is written into the correlation between each reversible takeover airspace to construct an airspace passage right migration matrix. The prevention and control party is defined as the initiator of the right-of-way migration action, and the drone operator is defined as the path connection response entity. A master-slave strategy space is established according to the airspace right-of-way migration matrix to obtain the strategy space dataset. A payoff function is constructed for the right-of-way migration action and path connection action in the policy space dataset. The payoff function is used to constrain the sequential switching relationship between the blockade occupation state, the detour transition state and the restoration release state, and a Stackelberg game model is established.

7. The intelligent airspace resource allocation method for low-altitude safety as described in claim 6, characterized in that, The steps for forming spatial game strategy information are as follows: Based on the Stackelberg game model, the payoff of right-of-way migration actions in the policy space dataset is evaluated according to the allowed switching order. Combinations of right-of-way migration actions that satisfy the sequential switching relationship are selected to determine the airspace partitioning strategy. The airspace partitioning strategy is input into the Stackelberg game model to limit the path connection response range of the drone operator, and the optimal path connection response action corresponding to the airspace partitioning strategy is solved according to the payoff function to determine the flight path strategy. The airspace division strategy and flight path strategy are combined and matched according to the reversible takeover airspace and corresponding time location, and then filtered according to the preset equilibrium conditions to form airspace game strategy information.

8. The intelligent airspace resource allocation method for low-altitude safety as described in claim 6 or 7, characterized in that, The steps for forming the airspace resource allocation scheme are as follows: The airspace division strategy in the airspace game strategy information is broken down into airspace occupation entries of each reversible takeover airspace at the corresponding time position, and the flight path strategy is broken down into path occupation entries of the flight target at the corresponding spatial position and the corresponding time position, forming a spatiotemporal occupation list. The spatial and temporal occupancy list is used to link and organize the mutual influence between airspace occupancy items and path occupancy items caused by the closed occupancy status, the transit status, and the resumption of passage status, forming a constraint write-back diagram. The mutual restraint relationships in the restraint and back-rewrite diagram are closed-loop merged, conflicting closed-loop clusters are identified, and blockade pruning, bypass insertion, and reset backfilling are performed to form an airspace resource allocation scheme.

9. The intelligent airspace resource allocation method for low-altitude safety as described in claim 8, characterized in that, The steps for continuously monitoring real-time flight status information via a 5G-A integrated sensing base station are as follows: The airspace resource allocation scheme is scheduled and parsed and converted into UAV flight scheduling and control commands to drive the UAV to execute flight missions and obtain actual flight execution status data. By using a 5G-A integrated sensing base station, the actual flight execution status data is continuously sensed and collected to form real-time flight status information.

10. The intelligent airspace resource allocation method for low-altitude safety as described in claim 1 or 9, characterized in that, The steps for generating a dynamic airspace resource scheduling report are as follows: The spatiotemporal deviation between real-time flight status information and flight path strategy in airspace resource allocation scheme is calculated to obtain airspace scheduling deviation index. Based on the airspace scheduling deviation index, the airspace resource allocation scheme is dynamically adjusted and the scheduling process information is recorded to form a dynamic airspace resource scheduling report.

Images