A highway unmanned aerial vehicle low-altitude intelligent traffic dynamic airspace management and control method

Abstract

The application belongs to the technical field of intelligent traffic and unmanned aerial vehicle management and control, and particularly relates to a highway unmanned aerial vehicle low-altitude intelligent traffic dynamic airspace management and control method. The method comprises the following steps: constructing a digital multi-dimensional space-time topology model of the highway low-altitude airspace, dividing the space into dynamic grid units and establishing a space-time coupling architecture; extracting unmanned aerial vehicle task characteristics and converting them into standardized task vectors; establishing a dynamic route right allocation mechanism based on a generalized second-price auction model, defining the right of way as a scarce resource for bidding; using an edge computing node to solve the Nash equilibrium solution in a distributed manner to determine the access authority and navigation priority of each unmanned aerial vehicle; generating a management and control instruction to realize accurate path guidance, and combining real-time flow state for closed-loop feedback regulation. The application combines game theory and auction mechanism in depth, improves resource allocation efficiency, ensures the priority of emergency tasks, and ensures the safety and system robustness of multi-machine concurrent operation.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation and unmanned aerial vehicle (UAV) management technology, specifically relating to a method for dynamic airspace management of low-altitude intelligent transportation for UAVs on highways. Background Technology

[0002] With the rapid evolution of the low-altitude economy and drone technology, their applications in highway inspection, emergency logistics, and intelligent traffic monitoring are becoming increasingly widespread, becoming crucial for building a modern, integrated transportation system. As a new type of production factor, the efficient utilization of low-altitude airspace is of great significance for improving the overall operational efficiency and safety of the transportation system. Especially along high-traffic highways, the concurrent operation of multiple drones poses challenges to the organization of airspace resources, access control, and dynamic coordination capabilities.

[0003] Dynamic airspace management technology aims to avoid flight conflicts and balance traffic flow in multi-mission scenarios by digitally modeling and allocating low-altitude airway resources. Existing management mechanisms typically rely on fixed physical intervals or predefined flight schedules, attempting to establish orderly flight corridors in three-dimensional space. This model is feasible to some extent in low-density operating environments, but with the explosive growth in the frequency of UAV operations, the problem between limited airspace resources and diversified flight demands is becoming increasingly prominent.

[0004] Traditional technical architectures heavily rely on a first-come, first-served linear scheduling logic, resulting in highly delayed and unfair airspace resource allocation, making it difficult to address conflicts of interest among multiple operators. Furthermore, existing solutions lack a deep understanding of the value and urgency of flight missions, preventing critical tasks such as medical rescue or accident handling from achieving priority passage in congested segments. In addition, low-altitude traffic systems incur enormous computational overhead when dealing with interdisciplinary game theory problems, making it difficult to achieve dynamic Nash equilibrium of airspace right-of-way through edge computing modules in a very short time, potentially leading to localized airspace paralysis. Summary of the Invention

[0005] The purpose of this invention is to provide a method for low-altitude intelligent traffic dynamic airspace management using unmanned aerial vehicles (UAVs) on highways, which can solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: a method for dynamic airspace management and control of low-altitude intelligent transportation by unmanned aerial vehicles on highways, comprising the following steps: constructing a digital multi-dimensional spatiotemporal topology model of the low-altitude airspace of highways, dividing the low-altitude three-dimensional space of the flight segment into multiple dynamic grid units with independent identifiers, and establishing physical associations and logical adjacency relationships between the grid units. The system acquires drone mission application information submitted by multiple operators in real time, identifies the mission's business type, expected flight path, scheduled time requirement, and mission attribute characteristics through a feature extraction module, and transforms the above information into a standardized mission vector. A dynamic right-of-way allocation mechanism based on the generalized second-price auction model is established. According to the urgency and value attributes in the standardized task vector, the logical bidding index of each task in the grid cell is calculated, and the airspace passage right is defined as a resource to be allocated. By utilizing edge computing nodes deployed along highways, distributed parallel computing is performed for multi-entity competition scenarios to solve the Nash equilibrium solution for airspace resource allocation, and to determine the access rights and flight priorities of each UAV within a time slice, i.e., the right-of-way allocation result. Dynamic airspace control instructions are generated based on the Nash equilibrium solution and sent to each UAV terminal through a low-altitude communication link to achieve flight path guidance and conflict avoidance. The system monitors real-time traffic flow in the airspace, compares the preset throughput threshold with the safety interval indicator, and uses closed-loop feedback adjustment logic to dynamically correct the right-of-way allocation results.

[0007] Preferably, the process of constructing a digital multidimensional spatiotemporal topology model of the low-altitude airspace of the highway includes: acquiring geographic information data of the entire highway, including the latitude and longitude coordinates of the road centerline, roadbed width, the location of guardrails along the route, and the clearance height of overpasses. Using the centerline of the highway as a reference, extend horizontally to both sides and vertically upwards to define a low-altitude flight corridor that closely follows the highway. In three-dimensional space, regular hexahedral voxels are used to discretize the low-altitude flight corridor, generating voxel units with unique spatial index numbers, wherein the side length of each voxel unit is set according to the complexity of the airspace. By introducing the time dimension as the fourth variable, the continuous time axis is divided into time steps of equal length, forming a spatiotemporally coupled four-dimensional grid architecture, so that each grid cell represents physical space and time slice. Meteorological environmental parameters, ground facility constraint data, and spatial distribution characteristics of electromagnetic interference intensity are embedded in the digital multidimensional spatiotemporal topology model in real time, and dynamic weight attributes are assigned to each grid cell. The dynamic weight attribute is composed of a traffic risk factor, an energy consumption compensation coefficient, and an airspace congestion level, which is used to reflect the accessibility, risk level, and expected energy consumption cost of the grid unit in a given time period.

[0008] Preferably, the processing of the drone mission application information includes: establishing a multi-criteria mission evaluation system, classifying the missions into categories such as public safety, medical rescue, infrastructure inspection, commercial logistics, and personal consumption; For different categories of tasks, extract their sensitivity coefficient to flight delay, tolerance range for track deviation, and risk cost of task failure; By using a normalization algorithm, multi-dimensional task features are mapped to a unified value space ranging from zero to one, generating weight factors that represent the comprehensive value of the task. The standardized task vector further includes the physical performance parameters of the UAV, including the climb rate, maximum horizontal cruising speed, endurance limit, and average latency characteristics of the onboard communication link under no-load and fully loaded conditions. The standardized task vector also includes a social benefit evaluation factor for the task, which is used to balance commercial interests and social public value in the game process, so that public welfare tasks have competitive weight when resources are limited.

[0009] Preferably, the dynamic right-of-way allocation mechanism based on the generalized second-price auction model includes: setting the right-of-way of the grid cell within a time period as the auction target; Each drone operator, as a bidder, automatically generates a bid price based on the standardized task vector through the airborne terminal or ground station system. The bid price reflects the virtual resource cost or priority points that the operator is willing to pay to obtain the right-of-way. In the bidding logic, all bidders submit their bids simultaneously within the same time window, and the system collects all bids and sorts them in descending order. The bidder ranked first in the bidding sequence obtains the right to pass through the corresponding grid cell within the target time period, and the final cost paid by the bidder is set as the second highest value in the bidding sequence. If there are only bidders in the bidding sequence, the bidders pay according to the preset base price, and the second price payment logic induces the bidders to bid according to the true value of the task.

[0010] Preferably, the process of solving the Nash equilibrium solution for spatial resource allocation includes: deploying edge processing units with vector processing capabilities at preset intervals along the highway, constructing a distributed computing network, and having each processing unit exchange data between nodes through a high-speed fiber optic backbone network to achieve global perception; A non-cooperative game model involving the decision-making behavior of multiple drones is established, and the strategy space of each drone is defined as a combination of the set of possible flight paths and the possible entry time window under the premise of satisfying physical constraints. A utility function is defined to quantify the merits of each strategy. The value of the utility function is equal to the preset total score minus the path length cost, time delay cost, auction payment cost, and collision risk cost. The path length cost is equal to the ratio of the actual flight distance to the shortest theoretical distance multiplied by a first preset ratio coefficient; the time delay cost is equal to the difference between the actual arrival time and the predetermined earliest arrival time multiplied by a second preset ratio coefficient; and the auction payment cost is equal to the ratio of the virtual payment amount to the initial virtual budget multiplied by a third preset ratio coefficient. By using an iterative search algorithm to find a steady state in a multidimensional policy space, the utility function value of any drone cannot be improved by unilaterally changing its flight path or speed, provided that the policies of other drones remain unchanged. For large-scale game problems with more participants than a preset number, a hierarchical and partitioned solution strategy is adopted. At the macro level, the traffic quota between different road segments is calculated, and at the micro level, right-of-way games are conducted within the grid cells.

[0011] Preferably, the process of generating dynamic airspace control instructions includes: converting the Nash equilibrium solution into specific control parameters, including the takeoff time of each UAV, the transit time of passing through latitude and longitude waypoints, the hard limits of the altitude level, and the allowable range of speed fluctuations in different flight segments; Establish redundant communication channels and use fifth-generation mobile communication technology or short-range communication technology to encapsulate the dynamic airspace control instructions into reliable data packets with check bits and encrypted signatures for transmission; Before the official instructions are issued, a conflict prediction and collision detection procedure is executed to verify the exclusivity of the generated trajectory in physical space. It is required that the distance between the center points of any two drones at any time is greater than three times the wingspan of the drone plus the safety margin. In response to signal jamming or link interruption, the emergency return path or coordinates of local hovering are pre-embedded in the dynamic airspace control command. If the UAV does not receive a subsequent heartbeat packet within a preset time, the emergency command is activated immediately.

[0012] Preferably, the process of monitoring the real-time traffic flow status in the airspace, comparing the preset throughput threshold with the safety interval index, and dynamically correcting the right-of-way allocation results using closed-loop feedback adjustment logic includes: using ground radar arrays deployed along the route, multispectral optoelectronic sensing equipment, and the positioning information fed back by UAVs through the Automatic Dependent Surveillance-Broadcast (ADS) protocol to obtain the actual three-dimensional coordinates and motion vectors of all operational targets in the airspace. The system continuously calculates the spatial and temporal deviations between the actual trajectory and the preset planned trajectory. When the spatial deviation exceeds the first preset spatial threshold or the temporal deviation exceeds the first preset temporal threshold, the local trajectory correction logic is triggered. The real-time traffic density within the statistical grid cell is calculated. When the real-time traffic density reaches a preset proportion of its physical carrying capacity limit, the area is determined to be in a potential congestion state and the auction bidding threshold coefficient of the area is automatically increased. By using a machine learning model based on long short-term memory networks, we can make rolling predictions of traffic trends over future time periods and adjust resource allocation weights in advance based on the prediction results. The system monitors the signal strength indicator and bit error rate in each grid cell in real time. When the signal reliability is lower than the preset reliability threshold, it automatically lowers the airspace carrying capacity limit of the area and forcibly increases the physical safety distance between drones.

[0013] Preferably, the method further includes a security verification and emergency avoidance step: after each round of auction and allocation, the system automatically performs a logical consistency check by scanning the temporal and spatial overlap of all allocated right-of-way to ensure that there is no right-of-way overlap or logical conflict. For urgent tasks of a certain level, the system activates the absolute priority passage protocol with the highest logical priority. The absolute priority passage protocol directly skips the auction bidding process and opens a green channel for the urgent task by forcibly requisitioning the allocated right-of-way. For other tasks affected by the expropriation of road rights, the system initiates social compensation calculation logic to automatically replenish virtual bid points or replan the optimal path in subsequent resource allocation rounds. For top-level emergency missions, all edge nodes along the route simultaneously issue full-channel broadcast guidance instructions, forcing non-emergency mission drones within the surrounding preset range to perform avoidance maneuvers such as descending in altitude or moving outwards from the flight path.

[0014] Preferably, the method further includes dynamic obstacle avoidance and node disaster recovery steps: integrating a dynamic obstacle avoidance layer into the digital multidimensional spatiotemporal topology model, and updating the temporary no-fly zone, sudden weather warning zone and ground construction area along the highway in real time by accessing an external information platform; The system dynamically reduces the passability weight of the corresponding grid cell based on the radius and duration of the obstacle's influence, so that the utility function in the game-solving phase will determine the behavior of entering the area as a low-utility state. The edge computing nodes use a peer-to-peer network architecture for information synchronization, and each node maintains a local topology copy and a task status table. When an edge node experiences a power outage or hardware failure, neighboring nodes detect the fault through a heartbeat mechanism and take over the control area of ​​the faulty node according to a preset jurisdiction takeover protocol. Each edge node's internal storage module records historical data of all drones within a specified range. This historical data includes flight trajectories, average energy consumption, and command execution accuracy. The historical data is then used to dynamically adjust the performance parameters in the game theory model.

[0015] Preferably, the method further includes air-ground collaborative management and blockchain evidence storage steps: real-time synchronization with the highway ground traffic flow center is performed through a standardized data exchange protocol. When a traffic accident or congestion occurs on a ground road section, the airspace layer above the congested road section is automatically adjusted and reconstructed into a ground auxiliary observation layer. The system guides drones that are performing inspection tasks nearby to converge on the congested area, and reallocates right-of-way through a game mechanism to ensure that drones with real-time video transmission capabilities are evenly distributed above the accident point. An air-ground collaborative utility term is introduced into the utility function of the right-of-way game. The magnitude of the air-ground collaborative utility term depends on whether the current observation perspective of the UAV can cover the blind spot of ground traffic. All bidding data from each round of auction participants, intermediate decision-making data, final execution trajectory instructions, and actual drone trajectory feedback data are recorded in a decentralized ledger database based on blockchain technology. Secure hashing algorithms and asymmetric cryptographic signatures are used to ensure the immutability of data after it is stored in the database, and a credit rating algorithm based on game theory behavior is established to dynamically adjust the credit score according to the accuracy of the operator's instruction execution.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. By introducing marginal game theory and a generalized second-price auction mechanism, this invention effectively solves the problem of unfair allocation of airspace resources under multi-party participation. It breaks away from the outdated logic of the traditional first-come, first-served approach, treats airspace passage rights as digital commodities, and uses economic models to guide resources toward high-value, high-urgency tasks, thereby improving the allocation efficiency of airspace resources from the underlying mechanism.

[0017] 2. The adoption of a distributed computing scheme based on Nash equilibrium significantly enhances the system's dynamic adjustment capabilities in complex traffic scenarios. Through parallel processing of edge computing nodes, multi-objective conflict avoidance and trajectory planning can be completed in a very short time, ensuring the real-time nature of control commands and preventing traffic paralysis in local airspace.

[0018] 3. The establishment of a digital multidimensional spatiotemporal topology model enables refined modeling of the low-altitude airspace of highways. By deeply coupling physical space, time dimension, and environmental constraints, a deterministic and predictable flight environment is provided for each UAV, improving system safety during multi-UAV concurrent operation.

[0019] 4. This invention achieves deep perception of mission value and differentiated services. Through standardized mission vectors, the system can identify and guarantee the absolute priority of critical missions such as medical rescue, enhancing social welfare benefits while providing a fair and competitive market environment for the operation of commercial drones.

[0020] 5. The introduction of a closed-loop feedback adjustment mechanism and safety verification steps constructs multiple lines of defense. Through real-time monitoring, deviation correction, and emergency protocols, the system ensures high robustness and reliability in the face of communication disturbances, hardware failures, or unexpected tasks, laying a solid technical foundation for the large-scale development of the low-altitude economy. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall technical solution architecture according to the present invention; Figure 2 This is a schematic diagram of the core principle framework of dynamic right-of-way allocation based on generalized second-price auction and Nash equilibrium game according to the present invention. Figure 3 This is a logical flowchart of the construction of a digital multidimensional spatiotemporal topology model and the standardization of task features according to the present invention. Figure 4 This is a schematic diagram illustrating the parallel game solving and multi-level interaction relationship based on distributed edge computing nodes according to the present invention; Figure 5 This is a flowchart illustrating the dynamic control command issuance and closed-loop flow monitoring feedback adjustment process according to the present invention. Detailed Implementation

[0022] Example 1: Please refer to the appendix Figure 1 To be continued Figure 5 To make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to specific embodiments.

[0023] In the dynamic airspace management method for low-altitude intelligent transportation of highway unmanned aerial vehicles implemented in this invention, the entire management process is divided into multiple deeply coupled logical stages, aiming to achieve precise scheduling of scarce airspace resources through the integration of economic game model and distributed computing architecture.

[0024] For step 1, a digital multi-dimensional spatiotemporal topology model of the low-altitude airspace of the highway is constructed. In this process, the system first acquires high-precision geographic information data of the entire highway, including the latitude and longitude coordinates of the road centerline, roadbed width, the location of guardrails along the route, and the clearance height of overpasses. Using the highway centerline as a reference, specific horizontal distances ranging from 50 to 150 meters are extended to both sides, and a specific vertical distance is extended upwards from ground level to a specific height of 300 meters, thereby defining the legal low-altitude flight corridor closely following the highway's direction.

[0025] In the 3D spatial modeling phase, the system discretizes the flight corridor using regular hexahedral voxels. The side length of each voxel is set between 3 and 10 meters depending on the complexity of the airspace, generating voxel units with unique spatial index numbers. To achieve fine-grained control, a time dimension is introduced as a fourth variable, dividing the continuous time axis into equal time steps of 100 milliseconds, forming a spatiotemporally coupled four-dimensional grid architecture. Each grid unit represents not only a specific physical space but also a specific time slice.

[0026] In the digital model, the system embeds meteorological and environmental parameters in real time, including instantaneous wind speed, visibility, and precipitation. Simultaneously, it integrates ground infrastructure constraint data, such as the location of high-voltage transmission lines, the distribution of base station towers, and the spatial distribution characteristics of electromagnetic interference intensity. The system assigns a dynamic weight attribute to each grid cell, which is composed of a traffic risk factor, an energy consumption compensation coefficient, and an airspace congestion level. This weight attribute reflects the accessibility, risk level, and expected energy consumption cost of a UAV flying in that area at a specific moment.

[0027] For step 2, the system acquires drone mission application information submitted by multiple operators in real time. The mission application process is implemented through a standardized application programming interface (API). The application data packet includes operator identification, drone hardware unique code, scheduled takeoff time, waypoint sequence, and mission urgency declaration. Through a feature extraction module, the system automatically identifies the mission's business type. Specifically, a multi-criteria mission evaluation system is established, classifying missions into five levels: public safety, medical rescue, infrastructure inspection, commercial logistics, and personal consumption. For different mission categories, a sensitivity coefficient to flight delay is extracted; for example, the sensitivity coefficient for medical rescue missions is set to the highest value. Simultaneously, the tolerance range for waypoint deviation is extracted, i.e., the maximum permissible distance the drone deviates from the predetermined centerline during mission execution.

[0028] Furthermore, the risk cost of mission failure needs to be extracted, i.e., the estimated social or economic loss if the mission cannot be completed on time. The system uses a normalization algorithm to map the aforementioned multi-dimensional mission features to a unified value space between 0 and 1, generating weighting factors representing the overall value of the mission. The standardized mission vector further includes the physical performance parameters of the UAV, specifically including climb rate, maximum horizontal cruising speed, maximum range limit, and average latency characteristics of the onboard communication link under empty and fully loaded conditions. In addition, the standardized mission vector also includes a social benefit evaluation factor for the mission, used to balance purely commercial interests and social public value in the game process, ensuring that public welfare missions have sufficient competitive weight when resources are limited.

[0029] For step 3, a dynamic right-of-way allocation mechanism based on a generalized second-price auction model is established. The system sets the right-of-way for airspace grid units within a specific time period of 1 to 5 seconds as the sole auction target. Each UAV operator acts as a bidder, and their onboard or ground station system automatically generates a bidding price based on the mission vector. This bidding price is not actual legal tender but reflects the virtual resource cost or priority points that the operator is willing to pay to acquire the right-of-way. In the bidding logic, each bidder submits a bid simultaneously within the same millisecond-level time window, and the system collects all bids and sorts them from highest to lowest. The bidder ranked first in the bidding sequence obtains the right-of-way for the corresponding grid unit within the target time period. The key economic logic is that the cost ultimately paid by the highest bidder is set as the second highest value in the bidding sequence. If there is only one bidder, the pre-set reserve price is paid. This mechanism induces bidders to bid according to the true value of the mission, effectively preventing malicious monopolization of airspace resources through significant price increases. Through this mechanism, airspace resources, as a scarce commodity, are allocated in a way that maximizes social welfare, ensuring that tasks with the most urgent airspace needs and the highest value contributions are given priority in obtaining right-of-way.

[0030] For step 4, the Nash equilibrium solution is solved using edge computing nodes deployed along the highway. High-performance vector processing units are deployed at predetermined intervals of 1 to 3 kilometers along the highway to construct a distributed computing network. Each processing unit is responsible for allocating airspace resources within its wireless coverage area, exchanging data between nodes through a high-speed fiber optic backbone to achieve global awareness of traffic flow status in adjacent road segments. During the solution process, a non-cooperative game model incorporating the decision-making behavior of multiple UAVs is established. The policy space of each UAV is defined as a combination of the set of possible flight paths and the possible entry time windows, satisfying physical constraints. A utility function is defined to quantify the merits of each policy. The utility function value equals the sum of the negative correlation terms of task completion time, energy consumption, collision risk probability, and the virtual cost paid in the auction mechanism. An iterative search algorithm is used to find a steady state in the multidimensional policy space.

[0031] Under this steady state, assuming that the strategies of all other drones remain unchanged, no single drone can improve its utility function value by unilaterally changing its flight path or speed, thus achieving Nash equilibrium. For large-scale game problems involving more than 100 participants, the system adopts a hierarchical and partitioned solution strategy. First, at the macro level, traffic quotas between different highway segments are calculated. Then, at the micro level, specific right-of-way games are conducted within each specific grid cell. Utilizing multi-level decoupling technology, the high-dimensional computational task is decomposed into multiple low-dimensional parallel subtasks, ensuring that the control scheme is output within a preset time threshold of 50 milliseconds.

[0032] For step 5, dynamic airspace control commands are generated based on the Nash equilibrium solution. The system transforms the abstract equilibrium solution into specific, executable control parameters, including the precise takeoff time of each UAV, the transit time through specific latitude and longitude waypoints, the hard limits of its altitude level, and the permissible speed fluctuation range within different flight segments. To ensure the reliability of command transmission, the system establishes redundant communication channels and utilizes 5G communication technology or dedicated short-range communication technology to encapsulate digital commands into reliable data packets with check bits and encrypted signatures for transmission. Before the commands are officially issued, the system executes conflict prediction and collision detection procedures to verify the exclusivity of the generated trajectories in physical space. This verification process requires that the distance between the center points of any two UAVs at any given time must be greater than three times the UAV's wingspan, plus a preset safety margin. In case of sudden signal jamming or link interruption, emergency return paths or hovering coordinates are pre-embedded in the commands. If a UAV does not receive a subsequent heartbeat packet within a preset time, these emergency commands will be activated immediately to ensure the overall robustness of the system.

[0033] For step 6, the system monitors real-time traffic flow within the airspace. Utilizing ground-based radar arrays deployed along the route, multispectral optoelectronic sensing equipment, and real-time positioning information fed back by UAVs via the ADS-B protocol, the system acquires the actual three-dimensional coordinates and motion vectors of all operational targets within the airspace. The system continuously calculates the spatial and temporal deviations between the actual flight paths and the pre-planned flight paths. When the spatial deviation exceeds 2 meters or the temporal deviation exceeds a first preset threshold of 500 milliseconds, the system immediately triggers local trajectory correction logic, guiding the UAVs back to their predetermined orbits by issuing instantaneous speed compensation commands. Simultaneously, the system calculates the real-time traffic density within specific grid cells, i.e., the number of UAVs per unit cubic space. When the density reaches 80% of its physical carrying capacity, a second preset threshold, the system determines that the area has entered a potential congestion state. At this point, the system automatically increases the auction bidding threshold coefficient for that area, guiding newly added tasks to choose detour routes or delay their entry through game theory algorithms. Furthermore, the system uses a machine learning model based on long short-term memory networks to perform rolling predictions of traffic trends over the next 15 minutes, adjusting resource allocation weights in advance.

[0034] Furthermore, the implementation of the above method also includes a security verification step. After each round of auction and allocation, the system automatically performs a logical consistency check. This check scans for temporal and spatial overlaps in all allocated right-of-way, ensuring that there are no bad debts due to overlapping right-of-way or logical conflicts. For specific levels of emergency tasks, such as organ transport or major traffic accident rescue, the system reserves an absolute priority passage protocol. This protocol has the highest logical priority and can directly bypass the auction bidding process. Once an emergency task triggers this protocol, the system will forcibly requisition the allocated right-of-way to create a clear green channel for the emergency task. For other tasks affected by the requisition of right-of-way, the system will automatically initiate compensation logic, providing them with priority compensation in subsequent auction rounds or replanning the optimal route for them free of charge.

[0035] The digital multidimensional spatiotemporal topology model also integrates a dynamic obstacle avoidance layer during implementation. This layer connects to an external public information platform to update in real time temporary no-fly zones, sudden weather warning zones, and ground construction areas along highways. Based on the obstacle's influence radius and duration, the system dynamically lowers the passability weight of the corresponding grid cell, and may even set its state to impassable. During the game-theoretic solution phase, the utility function classifies entering these high-cost areas as low-utility, inducing the algorithm to automatically avoid risky areas.

[0036] Collaboration between edge computing nodes employs a peer-to-peer network architecture for information synchronization. Each node maintains a local copy of the topology and a task status table. When an edge node fails due to a power outage or hardware malfunction, the two adjacent nodes quickly detect this through a heartbeat mechanism and, according to a pre-defined jurisdiction takeover protocol, each take over half of the failed node's control area. This decentralized disaster recovery mechanism ensures the continuity of control services. The internal storage modules of each node also record the historical flight trajectories, average energy consumption metrics, and command execution accuracy of all drones within a specific range. This historical data is used to dynamically adjust the performance parameters in the game theory model, improving prediction accuracy.

[0037] The right-of-way allocation results are ultimately issued in the form of a dynamic electronic fence. The UAV's onboard control system receives a set of boundary coordinates that change over time. Based on this boundary information, the onboard flight control system adjusts the motor speed and control surface deflection in real time to ensure that the UAV always operates within the allocated specific spatial tunnel. If the UAV illegally breaks through the electronic fence boundary due to external interference or other reasons, the control system will immediately trigger a three-level alarm logic. The first level is a voice and data link warning; the second level is a forced system intervention correction; and the third level is, in extremely dangerous situations, a forced takeover and guidance to the nearest emergency landing point.

[0038] To ensure the fairness and traceability of the control measures, this embodiment also involves data storage and traceability. The system records all auction participant bids, intermediate decision data, final execution trajectory instructions, and actual drone trajectory feedback in a decentralized ledger database based on blockchain technology. This database utilizes SHA-256 encryption technology and asymmetric encryption signatures to ensure that the data cannot be tampered with by any single entity after being stored. This provides objective and detailed data support for subsequent determinations of liability in traffic accidents, billing of right-of-way usage among different operators, and continuous iterative optimization of the control algorithm.

[0039] Regarding communication quality monitoring, the system monitors the signal strength indicator and bit error rate in each grid cell in real time. When the signal reliability in a specific area falls below a preset reliability threshold of 99%, the system automatically reduces the airspace carrying capacity limit for that area by 50%. In this way, the physical safety distance between drones is forcibly increased to compensate for the potential collision risk caused by increased communication latency. At the same time, the system will instruct the smart antenna arrays along the line to adjust the beam pointing, or dispatch dedicated communication relay drones to attempt to quickly restore the communication quality in that area.

[0040] The system also mandates that the drone's current remaining battery percentage and estimated energy consumption rate be included in the mission declaration information. In the utility function calculation of right-of-way game, the battery weight term increases exponentially as the remaining battery level decreases. This means that drones with low battery levels are more likely to be assigned the shortest, lowest-energy-consumption path, or the right-of-way near highway service area landing points, preventing crashes due to battery depletion.

[0041] Finally, this method achieves cross-domain collaboration with ground traffic flow control systems. By interfacing with the data interface of the highway monitoring center, when a major ground traffic accident causes traffic congestion, the system automatically senses the coordinates of the ground congestion and immediately releases more low-altitude airspace resources above the road segment, even temporarily converting the original logistics airspace into emergency command airspace. By dispatching multiple drones equipped with loudspeaker, supplementary lighting, and real-time mapping capabilities to the area, integrated air-ground traffic emergency guidance and refined management are achieved.

[0042] Example 2: Based on Example 1, Example 2 further describes in detail the specific micro-operation process of how edge computing nodes perform distributed parallel game solving, as well as the data scheduling logic in multi-machine high-concurrency scenarios.

[0043] In the specific implementation of step 4, each edge computing node is configured as an independent game processing unit. When a swarm of drones enters the jurisdiction of a specific edge node, that node first initiates a participant authentication procedure. The authentication process includes reading the drone's mission vector, extracting its mission level, total virtual budget, and expected entry / exit grid number.

[0044] The memory pool within the edge node is divided into multiple concurrent processing slots. Each slot is responsible for maintaining a local game model. For a new drone entering the jurisdiction, the edge node performs the following micro-operations: First, based on the drone's current position and target destination, it pre-selects 3 to 5 candidate paths in the digital multidimensional spatiotemporal topology model using a heuristic search algorithm. These candidate paths must completely avoid known static obstacles and dynamic no-fly zones. Next, for each candidate path, it calculates the set of grid cells for each time slice it traverses.

[0045] Next, the bidding process begins, including generation and sorting. Edge nodes simulate a generalized second-price auction, aggregating bids from all drones requesting the same grid cell within the same time step. In the bidding logic, the virtual bid amount equals the mission value weight multiplied by the urgency coefficient, plus the reciprocal of the remaining range multiplied by the safety weight coefficient. The sorting algorithm uses quicksort logic, generating a bid sequence within microseconds. The drone with the highest bid is pre-assigned the right-of-way to that grid, and its payment cost is set to the second highest value in the sequence.

[0046] After initially determining right-of-way allocation, the edge nodes initiate Nash equilibrium iterative verification. The system calculates the individual utility function value for each UAV under the current allocation scheme. The specific formula for calculating the utility function value is as follows: ; ; ; ; ; in, The individual utility value when choosing path p for the i-th drone; For path length cost, Let i be the actual flight distance of drone i on path p. Let i be the shortest theoretical distance from the starting point to the destination for drone i. As a result of time delay, Let $i$ be the actual arrival time of drone $i$ on path $p$. The earliest scheduled arrival time for drone i; To pay the price for the auction Let i be the virtual payment amount for drone i in obtaining path p. Let i be the initial virtual budget for drone i; As a consequence of collision risk, Let i be the density of neighboring drones on path p. This is a dynamic calculation function for collision risk and density.

[0047] If a drone discovers that switching to an alternative path can increase the utility function value by more than a preset small increment, the drone will update its policy. All participating entities repeat this process until no entity in the entire system is willing to unilaterally change its policy, or the number of iterations reaches the upper limit of 500. At this point, the system outputs the current policy combination as an approximate Nash equilibrium solution.

[0048] To improve computational efficiency, edge nodes employ a spatial partitioning-based data preprocessing technique. The jurisdictional area is divided into multiple sub-sectors, and data exchange only occurs between adjacent sub-sectors where trajectory intersections are possible. For physically isolated drone swarms, the system executes game theory solutions in parallel on different processor cores. This multi-level parallel architecture enables the system to support high-density management of over 200 drones per square kilometer.

[0049] During the command generation phase, the edge nodes decapsulate the Nash equalization into a specific binary control protocol format. Each control frame contains a 16-bit synchronization header, a 32-bit unique UAV identifier, a 64-bit spatiotemporal fence definition, a 32-bit velocity vector limit, and a 16-bit cyclic redundancy check (CRC) code. Commands are issued using asymmetric encryption to ensure that they are not intercepted or tampered with by third parties during transmission.

[0050] Example 3: This example 3 focuses on describing the automated response and resource reconfiguration mechanism of the method of the present invention in the context of extreme weather and sudden public safety incidents.

[0051] When meteorological monitoring sensors detect wind speeds exceeding 15 meters per second or visibility below 50 meters in a section of a highway, the digital multidimensional spatiotemporal topology model in step 1 will immediately trigger a real-time update. The system automatically calculates the impact of meteorological disasters on the flight stability of different types of drones. For example, for logistics drones with smaller rotor diameters, the system increases the weight of the passage cost in the affected area grid by 100 times, effectively creating a logical dynamic no-fly zone. For medium-sized industrial drones with higher wind resistance, the system only increases the weight by 2 times and forcibly limits their maximum flight speed to no more than 8 meters per second.

[0052] In step 2, during the task application phase, if the system receives a critical task instruction from the public security or medical departments, the priority bit in the standardized task vector will be set to the highest overflow state. At this point, the auction mechanism in step 3 will enter requisition mode. The system will no longer perform the usual price sorting, but will directly remove the grid cells required for the emergency task from the auction pool and lock them for special purposes.

[0053] In the game-theoretic solution of step 4, the system initiates social compensation calculation logic for other affected business tasks. Because the emergency task requisitioned existing right-of-way, the system, which was originally in Nash equilibrium, is disrupted. Edge computing nodes quickly identify which drones' utility functions are negatively affected and automatically replenish virtual bid points for these affected drones in the next round of resource allocation. For example, if a drone is delayed by 5 minutes due to avoiding a rescue mission, the system will compensate it with credits equivalent to 30% of the mission's value in its virtual account, ensuring it gains a compensatory advantage in subsequent right-of-way competition.

[0054] In step 5, during the command issuance phase, for critical missions, the system will initiate full-channel broadcast guidance. All edge nodes along the route will simultaneously issue airspace clearing commands, forcing non-emergency mission drones within a 500-meter radius to either descend or move outwards to avoid the obstacle. The execution of these avoidance commands is monitored in real-time by the system in step 6. If a commercial drone fails to respond within 3 seconds, the system will transmit directional jamming signals via a ground base station, triggering the underlying security protection program on the airborne device and forcing it into hover mode until the emergency mission is completed.

[0055] Furthermore, this embodiment details the application of data storage and traceability in accident handling. Assuming a collision occurs between two aircraft under complex weather conditions, the accident investigation system will retrieve all relevant auction data and Nash equilibrium solutions from the decentralized ledger database within the five minutes prior to the accident. By comparing the preset command trajectory, the actual feedback trajectory, and the decision logic of edge nodes, the system can automatically determine responsibility for the accident. For example, if data records show that a certain operator's drone, after receiving a deceleration command, failed to reduce its speed vector within the predetermined time, the system will determine that a malfunction in the drone's flight control system or a violation of regulations by the operator is the primary cause of the accident. This automatic determination mechanism based on an immutable ledger reduces the administrative costs of low-altitude traffic management.

[0056] Example 4: This example 4 details the specific implementation path of air-ground collaborative management and control in the method of the present invention, as well as the deep integration logic with the ground intelligent transportation system.

[0057] During the monitoring process in step 6, this system synchronizes data in real time with the highway surface traffic flow center through a standardized data exchange protocol. When the surface road administration system detects a serious traffic accident or a road congestion exceeding 2 kilometers in length, this information is converted into spatial constraint variables and fed back into the topology model in step 1.

[0058] The system automatically adjusts the airspace layering above congested road sections. Lower altitude layers originally used for commercial express delivery (e.g., 30 to 60 meters above ground) are temporarily reconfigured as ground-based auxiliary observation layers. The system guides drones performing inspection tasks nearby to converge on this area and reallocates right-of-way through a game-theoretic mechanism to ensure that these drones, equipped with real-time video transmission capabilities, are evenly distributed above the accident point in a circling obstacle avoidance manner.

[0059] During the mission declaration phase, the system identifies a special type of collaborative mission. The standardized vector for this type of mission includes ground traffic condition feedback parameters. When a drone performs this type of mission, its virtual budget during the game process is directly subsidized by the right-of-way management system, without consuming the operator's own budget.

[0060] In the Nash equilibrium solution in step 4, the system introduces an air-ground collaborative utility term. The magnitude of this term depends on whether the UAV's current observation perspective can cover the blind spots of ground traffic. If a UAV's flight strategy can simultaneously consider its own transportation mission and the ground traffic observation needs, its utility function will receive an additional gain reward. This mechanism encourages UAVs to actively participate in the overall governance of intelligent transportation while completing their own tasks.

[0061] In the control instructions generated in step 5, the system adds task overlap instructions for air-to-ground collaborative scenarios. For example, it requires the drone to automatically lock its gimbal camera onto a specific direction of the highway when passing through a certain road segment. Simultaneously, through the closed-loop feedback in step 6, the system dynamically adjusts the open ratio of airspace resources based on the progress of ground traffic management. Once the incident is resolved and traffic flow is restored, the system automatically removes the ground-based auxiliary observation layer, returning the airspace to the regular commercial auction logic, thus achieving flexible adjustment of airspace resource utilization in cross-dimensional traffic scenarios.

[0062] Example 5: This Example 5 describes in detail the dynamic relay control logic of UAV energy efficiency management and long-range missions in the method of the present invention.

[0063] In step 2, the mission declaration phase, the standardized mission vector strictly defines the UAV's energy model. This model includes the total battery capacity, current remaining power, average energy consumption rate per unit distance, and energy consumption variation curves at different flight speeds. The system requires all declared missions to reserve 20% of the legally mandated reserve power.

[0064] In the auction mechanism of step 3, a continuous right-of-way pre-auction mode is introduced for long-range missions. Since long-range missions involve crossing multiple edge node jurisdictions, the system automatically calculates all grid cells the mission will traverse within the next 30 minutes. Bidders can submit a set of related bids, and the system uses a predictive algorithm to assess the competitive pressure on subsequent road segments. If severe congestion is anticipated on subsequent road segments, the system will guide the drone to choose a slightly longer but more energy-efficient and less competitive path in the current round.

[0065] In the game-theoretic solution process in step 4, a power decay risk term is added to the utility function. The value of this term is inversely proportional to the drone's remaining power. When a drone's power drops below 30%, its power decay risk term increases rapidly, forcing the game outcome to favor that drone, allowing it to pass through the jurisdiction with minimal maneuvering and the most constant speed. If the calculation finds that no game strategy can guarantee the drone's safe arrival at its destination, the edge node will immediately generate an emergency landing command.

[0066] In the instructions generated in step 5, the system allocates a dedicated energy-saving flight path for this type of low-power drone. This flight path is located at the edge of the flight corridor, away from major traffic flows, and its altitude level is set to the range most conducive to maintaining a constant speed. Simultaneously, the system uses a low-altitude communication link to query the occupancy status of automatic drone battery swapping stations or charging piles in highway service areas along the route in real time.

[0067] In step 6, during monitoring and feedback, the system continuously compares the drone's theoretically estimated remaining battery power with the actual remaining battery power. If the deviation exceeds 5%, the system will re-execute steps 3 and 4, adjusting subsequent right-of-way in real time. For example, if headwinds cause rapid battery depletion, the system will automatically increase its priority in subsequent auctions by one level, ensuring it no longer performs energy-intensive hovering maneuvers while waiting for allocation. Through this dynamic game-theoretic mechanism based on energy efficiency awareness, this invention significantly reduces the risk of drone crashes in highway airspace, ensuring the reliability of long-haul drone logistics.

[0068] Example 6: This Example 6 describes in detail the refined flow control logic of the method of the present invention for large-scale drone swarms in hub interchange areas.

[0069] Highway interchanges are core areas of airspace convergence, and their spatiotemporal topology model (step 1) is refined into irregular voxel meshes with higher resolution in these areas. The voxel side length is reduced from 5 meters to 2 meters. Computationally enhanced edge nodes are deployed at the center points of the interchanges.

[0070] In step 2, task vectors traversing interchanges are marked as hub crossing patterns. These tasks have extremely low tolerance for track deviations. In the auction of step 3, grid cells above interchanges are designated as high-value targets, with their bidding frequency increased from once per second to once every 200 milliseconds.

[0071] In the game theory solution of step 4, a spatial decoupling and time synchronization strategy is adopted for the interchange area. Edge nodes divide the interchange airspace into 6 vertical height levels. The game model will prioritize guiding drones with different directions to different height levels. In the extreme case where multiple routes intersect at the same height level, the Nash equilibrium solver will forcibly introduce a time slot offset strategy, that is, by fine-tuning the passage time of each drone, it will achieve staggered passage in physical space to avoid the risk of collision.

[0072] In step 5, during the instruction issuance, the control system will issue a trajectory locking packet in real time for drones located in the interchange area. This data packet contains extremely high-frequency positioning correction parameters, requiring the drone navigation system to perform position calibration every 100 milliseconds.

[0073] For step 6, an encrypted millimeter-wave radar network was deployed in the interchange area for comprehensive trajectory tracking of UAVs. If monitoring detects that a UAV is not following the preset Nash equilibrium path, the system will immediately activate a global avoidance protocol, notifying all other UAVs in the interchange area to perform emergency collision avoidance maneuvers. This enhanced control mode for hub areas ensures that the airspace system can maintain efficient and orderly operation even at complex flight path nodes.

[0074] Example 7: This Example 7 describes in detail how the method of the present invention utilizes the decentralized ledger of blockchain to achieve automated right-of-way clearing and credit evaluation among operators.

[0075] In the data storage and traceability phase of the method, the virtual price of each auction is recorded in the blockchain transaction entry. At the end of the month or within a specific settlement cycle, this virtual price is converted into service fees or credit lines between operators based on a preset exchange rate.

[0076] The system establishes a credit rating algorithm based on game theory behavior. If an operator's drones consistently execute the instructions generated by the Nash equilibrium solution in multiple missions, and their feedback trajectories closely match the planned trajectories, the system will automatically improve the operator's credit score. In the auction mechanism of step 3, operators with high credit scores can obtain a certain percentage of bid discount or initial priority bonus when bidding for the same right-of-way.

[0077] Conversely, if an operator's drones frequently deviate illegally, hover without cause, or respond slowly during emergency maneuvers, the system will reduce their credit score in real time and record the violation in the ledger. In serious cases, their mission applications in step 2 will be automatically rejected by the system, or they will face extremely high bidding thresholds in step 3.

[0078] This decentralized ledger-based credit and settlement mechanism has built a self-regulating ecosystem. To reduce operating costs and improve the success rate of obtaining right-of-way, operators proactively upgrade their flight control algorithms and strengthen the maintenance of their drones. This, from a market competition perspective, drives improvements in the safety and efficiency of the entire low-altitude intelligent transportation system.

[0079] In summary, this invention achieves refined airspace modeling through a digital multidimensional spatiotemporal topology model, realizes fair and efficient resource allocation using a generalized second-price auction mechanism, solves the conflict avoidance problem under multi-stakeholder competition through distributed edge computing based on Nash equilibrium, and constructs a closed-loop security system by combining real-time monitoring and feedback adjustment. By introducing air-ground collaboration, energy efficiency management, hub flow control, and blockchain credit mechanisms, this invention provides a complete, robust, and scalable technical framework for the development of the low-altitude economy along highways.

[0080] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for dynamic airspace management and control of low-altitude intelligent transportation using unmanned aerial vehicles (UAVs) on highways, characterized in that: Includes the following steps: A digital multidimensional spatiotemporal topology model of the low-altitude airspace of highways is constructed, dividing the low-altitude three-dimensional space of the flight segment into multiple dynamic grid units with independent identification, and establishing physical associations and logical adjacency relationships between the grid units. The system acquires drone mission application information submitted by multiple operators in real time, identifies the mission's business type, expected flight path, scheduled time requirement, and mission attribute characteristics through a feature extraction module, and transforms the above information into a standardized mission vector. A dynamic right-of-way allocation mechanism based on the generalized second-price auction model is established. According to the urgency and value attributes in the standardized task vector, the logical bidding index of each task in the grid cell is calculated, and the airspace passage right is defined as a resource to be allocated. By utilizing edge computing nodes deployed along highways, distributed parallel computing is performed for multi-entity competition scenarios to solve the Nash equilibrium solution for airspace resource allocation, and to determine the access rights and flight priorities of each UAV within a time slice, i.e., the right-of-way allocation result. Dynamic airspace control instructions are generated based on the Nash equilibrium solution and sent to each UAV terminal through a low-altitude communication link to achieve flight path guidance and conflict avoidance. The system monitors real-time traffic flow in the airspace, compares the preset throughput threshold with the safety interval indicator, and uses closed-loop feedback adjustment logic to dynamically correct the right-of-way allocation results.

2. The method for dynamic airspace management of low-altitude intelligent transportation using unmanned aerial vehicles (UAVs) on highways according to claim 1, characterized in that, The process of constructing a digital multidimensional spatiotemporal topology model of the low-altitude airspace of the highway includes: acquiring geographic information data of the entire highway, including the latitude and longitude coordinates of the road centerline, roadbed width, the location of guardrails along the route, and the clearance height of overpasses. Using the centerline of the highway as a reference, extend horizontally to both sides and vertically upwards to define a low-altitude flight corridor that closely follows the highway. In three-dimensional space, regular hexahedral voxels are used to discretize the low-altitude flight corridor, generating voxel units with unique spatial index numbers, wherein the side length of each voxel unit is set according to the complexity of the airspace. By introducing the time dimension as the fourth variable, the continuous time axis is divided into time steps of equal length, forming a spatiotemporally coupled four-dimensional grid architecture, so that each grid cell represents physical space and time slice. Meteorological environmental parameters, ground facility constraint data, and spatial distribution characteristics of electromagnetic interference intensity are embedded in the digital multidimensional spatiotemporal topology model in real time, and dynamic weight attributes are assigned to each grid cell. The dynamic weight attribute is composed of a traffic risk factor, an energy consumption compensation coefficient, and an airspace congestion level, which is used to reflect the accessibility, risk level, and expected energy consumption cost of the grid unit in a given time period.

3. The method for dynamic airspace management of low-altitude intelligent transportation using unmanned aerial vehicles (UAVs) on highways according to claim 1, characterized in that, The processing of the drone mission application information includes: establishing a multi-criteria mission evaluation system, classifying the missions into categories such as public safety, medical rescue, infrastructure inspection, commercial logistics, and personal consumption; For different categories of tasks, extract their sensitivity coefficient to flight delay, tolerance range for track deviation, and risk cost of task failure; By using a normalization algorithm, multi-dimensional task features are mapped to a unified value space ranging from zero to one, generating weight factors that represent the comprehensive value of the task. The standardized task vector further includes the physical performance parameters of the UAV, including the climb rate, maximum horizontal cruising speed, endurance limit, and average latency characteristics of the onboard communication link under no-load and fully loaded conditions. The standardized task vector also includes a social benefit evaluation factor for the task, which is used to balance commercial interests and social public value in the game process, so that public welfare tasks have competitive weight when resources are limited.

4. The method for dynamic airspace management of low-altitude intelligent transportation using unmanned aerial vehicles (UAVs) on highways according to claim 1, characterized in that, The dynamic right-of-way allocation mechanism based on the generalized second-price auction model includes: setting the right-of-way of the grid unit within a time period as the auction target; Each drone operator, as a bidder, automatically generates a bid price based on the standardized task vector through the airborne terminal or ground station system. The bid price reflects the virtual resource cost or priority points that the operator is willing to pay to obtain the right-of-way. In the bidding logic, all bidders submit their bids simultaneously within the same time window, and the system collects all bids and sorts them in descending order. The bidder ranked first in the bidding sequence obtains the right to pass through the corresponding grid cell within the target time period, and the final cost paid by the bidder is set as the second highest value in the bidding sequence. If there are only bidders in the bidding sequence, the bidders pay according to the preset base price, and the second price payment logic induces the bidders to bid according to the true value of the task.

5. The method for dynamic airspace management of low-altitude intelligent transportation using unmanned aerial vehicles (UAVs) on highways according to claim 1, characterized in that, The process of solving the Nash equilibrium solution for spatial resource allocation includes: deploying edge processing units with vector processing capabilities at preset intervals along the highway, constructing a distributed computing network, and exchanging data between nodes through a high-speed fiber optic backbone network to achieve global perception. A non-cooperative game model involving the decision-making behavior of multiple drones is established, and the strategy space of each drone is defined as a combination of the set of possible flight paths and the possible entry time window under the premise of satisfying physical constraints. A utility function is defined to quantify the merits of each strategy. The value of the utility function is equal to the preset total score minus the path length cost, time delay cost, auction payment cost, and collision risk cost. The path length cost is equal to the ratio of the actual flight distance to the shortest theoretical distance multiplied by a first preset ratio coefficient; the time delay cost is equal to the difference between the actual arrival time and the predetermined earliest arrival time multiplied by a second preset ratio coefficient; and the auction payment cost is equal to the ratio of the virtual payment amount to the initial virtual budget multiplied by a third preset ratio coefficient. By using an iterative search algorithm to find a steady state in a multidimensional policy space, the utility function value of any drone cannot be improved by unilaterally changing its flight path or speed, provided that the policies of other drones remain unchanged. For large-scale game problems with more participants than a preset number, a hierarchical and partitioned solution strategy is adopted. At the macro level, the traffic quota between different road segments is calculated, and at the micro level, right-of-way games are conducted within the grid cells.

6. The method for dynamic airspace management of low-altitude intelligent transportation using unmanned aerial vehicles (UAVs) on highways according to claim 1, characterized in that, The process of generating dynamic airspace control instructions includes: converting the Nash equilibrium solution into specific control parameters, including the takeoff time of each UAV, the transit time of passing through latitude and longitude waypoints, the hard limits of the altitude level, and the allowable range of speed fluctuations in different flight segments; Establish redundant communication channels and use fifth-generation mobile communication technology or short-range communication technology to encapsulate the dynamic airspace control instructions into reliable data packets with check bits and encrypted signatures for transmission; Before the official instructions are issued, a conflict prediction and collision detection procedure is executed to verify the exclusivity of the generated trajectory in physical space. It is required that the distance between the center points of any two drones at any time is greater than three times the wingspan of the drone plus the safety margin. In response to signal jamming or link interruption, the emergency return path or coordinates of local hovering are pre-embedded in the dynamic airspace control command. If the UAV does not receive a subsequent heartbeat packet within a preset time, the emergency command is activated immediately.

7. The method for dynamic airspace management of low-altitude intelligent transportation using unmanned aerial vehicles (UAVs) on highways according to claim 1, characterized in that, The process of monitoring real-time traffic flow in the airspace, comparing the preset throughput threshold with the safety interval index, and dynamically correcting the right-of-way allocation results using closed-loop feedback adjustment logic includes: using ground radar arrays deployed along the route, multispectral optoelectronic sensing equipment, and the positioning information fed back by UAVs through the Automatic Dependent Surveillance-Broadcast (ADS) protocol to obtain the actual three-dimensional coordinates and motion vectors of all operational targets in the airspace. The system continuously calculates the spatial and temporal deviations between the actual trajectory and the preset planned trajectory. When the spatial deviation exceeds the first preset spatial threshold or the temporal deviation exceeds the first preset temporal threshold, the local trajectory correction logic is triggered. The real-time traffic density within the statistical grid cell is calculated. When the real-time traffic density reaches a preset proportion of its physical carrying capacity limit, the area is determined to be in a potential congestion state and the auction bidding threshold coefficient of the area is automatically increased. By using a machine learning model based on long short-term memory networks, we can make rolling predictions of traffic trends over future time periods and adjust resource allocation weights in advance based on the prediction results. The system monitors the signal strength indicator and bit error rate in each grid cell in real time. When the signal reliability is lower than the preset reliability threshold, it automatically lowers the airspace carrying capacity limit of the area and forcibly increases the physical safety distance between drones.

8. The method for dynamic airspace management of low-altitude intelligent transportation using unmanned aerial vehicles (UAVs) on highways according to claim 1, characterized in that, The method also includes security verification and emergency avoidance steps: after each round of auction and allocation, the system automatically performs a logical consistency check by scanning the temporal and spatial overlap of all allocated right-of-way to ensure that there is no right-of-way overlap or logical conflict. For urgent tasks of a certain level, the system activates the absolute priority passage protocol with the highest logical priority. The absolute priority passage protocol directly skips the auction bidding process and opens a green channel for the urgent task by forcibly requisitioning the allocated right-of-way. For other tasks affected by the expropriation of road rights, the system initiates social compensation calculation logic to automatically replenish virtual bid points or replan the optimal path in subsequent resource allocation rounds. For top-level emergency missions, all edge nodes along the route simultaneously issue full-channel broadcast guidance instructions, forcing non-emergency mission drones within the surrounding preset range to perform avoidance maneuvers such as descending in altitude or moving outwards from the flight path.

9. A method for dynamic airspace management of low-altitude intelligent transportation using unmanned aerial vehicles (UAVs) on highways according to claim 1, characterized in that, The method also includes dynamic obstacle avoidance and node disaster recovery steps: a dynamic obstacle avoidance layer is integrated into the digital multidimensional spatiotemporal topology model, and temporary no-fly zones, sudden weather warning zones and ground construction areas along the highway are updated in real time by accessing an external information platform; The system dynamically reduces the passability weight of the corresponding grid cell based on the radius and duration of the obstacle's influence, so that the utility function in the game-solving phase will determine the behavior of entering the area as a low-utility state. The edge computing nodes use a peer-to-peer network architecture for information synchronization, and each node maintains a local topology copy and a task status table. When an edge node experiences a power outage or hardware failure, neighboring nodes detect the fault through a heartbeat mechanism and take over the control area of ​​the faulty node according to a preset jurisdiction takeover protocol. Each edge node's internal storage module records historical data of all drones within a specified range. This historical data includes flight trajectories, average energy consumption, and command execution accuracy. The historical data is then used to dynamically adjust the performance parameters in the game theory model.

10. A method for dynamic airspace management of low-altitude intelligent transportation using unmanned aerial vehicles (UAVs) on highways according to claim 1, characterized in that, The method also includes air-ground collaborative management and blockchain evidence storage steps: real-time synchronization with the highway ground traffic flow center is achieved through a standardized data exchange protocol. When a traffic accident or congestion occurs on a ground road section, the airspace layer above the congested road section is automatically adjusted and reconstructed into a ground auxiliary observation layer. The system guides drones that are performing inspection tasks nearby to converge on the congested area, and reallocates right-of-way through a game mechanism to ensure that drones with real-time video transmission capabilities are evenly distributed above the accident point. An air-ground collaborative utility term is introduced into the utility function of the right-of-way game. The magnitude of the air-ground collaborative utility term depends on whether the current observation perspective of the UAV can cover the blind spot of ground traffic. All bidding data from each round of auction participants, intermediate decision-making data, final execution trajectory instructions, and actual drone trajectory feedback data are recorded in a decentralized ledger database based on blockchain technology. Secure hashing algorithms and asymmetric cryptographic signatures are used to ensure the immutability of data after it is stored in the database, and a credit rating algorithm based on game theory behavior is established to dynamically adjust the credit score according to the accuracy of the operator's instruction execution.

Images