Low-altitude airspace resource allocation method and device, storage medium and electronic equipment

Abstract

The invention provides a low-altitude airspace resource allocation method and device, a storage medium and electronic equipment, and is applied to the technical field of artificial intelligence. According to the method, the operation data of the unmanned aerial vehicle, the flight task queuing state and the historical task flow data are collected in real time, the current low-altitude airspace resource state diagram and the flight task demand prediction heat map are established, and the future task request can be accurately predicted. After the information is input into a low-altitude airspace resource allocation model, an unmanned aerial vehicle scheduling strategy can be optimized, the response speed and efficiency of resource allocation are improved, and airspace conflicts and resource waste are reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, and in particular to a low-altitude airspace resource configuration method and device, a storage medium and an electronic device. BACKGROUND

[0002] In the current management and scheduling of low-altitude airspace resources, traditional methods mainly rely on static division rules and manual instruction control.

[0003] However, although these traditional methods have a certain feasibility in early airspace management, with the increasing demand for low-altitude flights, such as logistics drones, electric vertical takeoff and landing (eVTOL) and urban air mobility (UAM), this management method gradually exposes its problems of reaction lag and insufficient efficiency. In the face of high-density, dynamic and multi-target resource configuration requirements, existing technologies are difficult to adapt to new challenges.

[0004] Therefore, how to improve the efficiency of low-altitude airspace resource configuration has become a technical problem that needs to be solved by those skilled in the art. SUMMARY

[0005] In view of the above problems, the present application provides a low-altitude airspace resource configuration method, device, storage medium and electronic device to overcome the above problems or at least partially solve the above problems, and the technical solutions are as follows:

[0006] A low-altitude airspace resource configuration method comprises:

[0007] Obtaining real-time unmanned aerial vehicle operation data, real-time flight task queuing state and historical task flow data of a target city low-altitude airspace, wherein the real-time unmanned aerial vehicle operation data includes real-time unmanned aerial vehicle flight data, airspace map information, scheduling history data, current external variable data and real-time flight request data of the target city low-altitude airspace at the current time, and the historical task flow data includes historical external variable data and historical flight request data of the target city low-altitude airspace at at least one historical time;

[0008] Using the real-time unmanned aerial vehicle operation data, obtaining a current low-altitude airspace resource state graph of the target city low-altitude airspace, wherein the current low-altitude airspace resource state graph includes airspace feature vectors of a plurality of airspace units in an airspace adjacency graph constructed based on the target city low-altitude airspace;

[0009] inputting the historical task flow data into a pre-constructed flight demand prediction model to obtain a flight task demand prediction heat map of the target city low-altitude airspace output by the flight demand prediction model, wherein the flight task demand prediction heat map comprises a task request prediction result of each airspace unit in a future period of time;

[0010] inputting the real-time flight task queuing state, the current low-altitude airspace resource state map, and the flight task demand prediction heat map into a pre-constructed low-altitude airspace resource configuration model to obtain a UAV scheduling strategy output by the low-altitude airspace resource configuration model.

[0011] Optionally, after the real-time flight task queuing state, the current low-altitude airspace resource state map, and the flight task demand prediction heat map are input into the pre-constructed low-altitude airspace resource configuration model to obtain the UAV scheduling strategy output by the low-altitude airspace resource configuration model, the method further comprises:

[0012] issuing the UAV scheduling strategy to a UAV operation platform, so that the UAV operation platform performs a UAV flight task in the target city low-altitude airspace according to the UAV scheduling strategy.

[0013] Optionally, after the UAV scheduling strategy is issued to the UAV operation platform, so that the UAV operation platform performs a UAV flight task in the target city low-altitude airspace according to the UAV scheduling strategy, the method further comprises:

[0014] obtaining actual scheduling execution data of the UAV flight task, wherein the actual scheduling execution data comprises a task completion condition, a flight conflict rate, and a delay quantity of the UAV flight task;

[0015] calculating evaluation index detection data of the UAV scheduling strategy based on the actual scheduling execution data, wherein the evaluation index detection data comprises an airspace utilization rate, a user satisfaction score, and a system stability;

[0016] updating a reward function of the low-altitude airspace resource configuration model based on the evaluation index detection data to obtain an updated low-altitude airspace resource configuration model.

[0017] Optionally, before the current low-altitude airspace resource state map of the target city low-altitude airspace is obtained by using the real-time UAV operation data, the method further comprises:

[0018] performing missing value filling and data standardization processing on the real-time UAV operation data.

[0019] Optionally, the airspace adjacency graph is obtained by modeling airspace states of the target city low-altitude airspace based on a graph neural network, wherein nodes in the airspace adjacency graph represent the airspace units, and edges in the airspace adjacency graph represent physical or scheduling connections between two adjacent airspace units.

[0020] Optionally, the flight demand prediction model is constructed based on a deep learning neural network and improves prediction spatial correlation through a spatio-temporal graph convolution network.

[0021] Optionally, the low-altitude airspace resource configuration model is constructed using a proximal policy optimization algorithm, and generates the unmanned aerial vehicle scheduling strategy based on a reward function R = λ1U - λ2C - λ3D, wherein U is resource utilization efficiency, C is airspace conflict quantity, D is average task delay time, and λ1, λ2 and λ3 are adjustable weight coefficients.

[0022] A low-altitude airspace resource configuration device comprises a data acquisition unit, a current low-altitude airspace resource state graph obtaining unit, a flight task demand prediction heat map obtaining unit, and an unmanned aerial vehicle scheduling strategy obtaining unit,

[0023] The data acquisition unit is configured to obtain real-time unmanned aerial vehicle operation data, real-time flight task queuing states, and historical task flow data of a target city low-altitude airspace, wherein the real-time unmanned aerial vehicle operation data comprises real-time unmanned aerial vehicle flight data, airspace map information, scheduling history data, current external variable data, and real-time flight request data of the target city low-altitude airspace at a current time, and the historical task flow data comprises historical external variable data and historical flight request data of the target city low-altitude airspace at at least one historical time;

[0024] The current low-altitude airspace resource state graph obtaining unit is configured to obtain a current low-altitude airspace resource state graph of the target city low-altitude airspace by using the real-time unmanned aerial vehicle operation data, wherein the current low-altitude airspace resource state graph comprises airspace feature vectors of a plurality of airspace units in an airspace adjacency graph constructed based on the target city low-altitude airspace;

[0025] The flight task demand prediction heat map obtaining unit is configured to input the historical task flow data into a pre-constructed flight demand prediction model, and obtain a flight task demand prediction heat map of the target city low-altitude airspace output by the flight demand prediction model, wherein the flight task demand prediction heat map comprises task request prediction results of each airspace unit within a future period of time;

[0026] The unmanned aerial vehicle scheduling strategy obtaining unit is configured to input the real-time flight task queuing state, the current low-altitude airspace resource state graph, and the flight task demand prediction heat map into a pre-constructed low-altitude airspace resource configuration model to obtain an unmanned aerial vehicle scheduling strategy output by the low-altitude airspace resource configuration model.

[0027] A computer-readable storage medium having a program stored thereon, wherein the program, when executed by a processor, implements the low-altitude airspace resource configuration method.

[0028] An electronic device includes at least one processor and at least one memory connected to the processor, wherein the processor and the memory complete communication with each other through a bus; the processor is configured to invoke program instructions in the memory to execute the low-altitude airspace resource configuration method.

[0029] According to the technical scheme, the low-altitude airspace resource configuration method, device, storage medium, and electronic device are provided, real-time unmanned aerial vehicle operation data, real-time flight task queuing state, and historical task flow data of a target city low-altitude airspace are obtained, the real-time unmanned aerial vehicle operation data includes real-time unmanned aerial vehicle flight data, airspace map information, scheduling historical data, current external variable data, and real-time flight request data of the target city low-altitude airspace at a current time, and the historical task flow data includes historical external variable data and historical flight request data of the target city low-altitude airspace at at least one historical time; the real-time unmanned aerial vehicle operation data is used to obtain a current low-altitude airspace resource state graph of the target city low-altitude airspace, the current low-altitude airspace resource state graph includes airspace feature vectors of a plurality of airspace units in an airspace adjacency graph constructed based on the target city low-altitude airspace; the historical task flow data is input into a pre-constructed flight demand prediction model to obtain a flight task demand prediction heat map of the target city low-altitude airspace output by the flight demand prediction model, the flight task demand prediction heat map includes task request prediction results of each airspace unit in a future period of time; the real-time flight task queuing state, the current low-altitude airspace resource state graph, and the flight task demand prediction heat map are input into a pre-constructed low-altitude airspace resource configuration model to obtain an unmanned aerial vehicle scheduling strategy output by the low-altitude airspace resource configuration model. The real-time unmanned aerial vehicle operation data, flight task queuing state, and historical task flow data are collected in real time, the current low-altitude airspace resource state graph and the flight task demand prediction heat map are established, and accurate prediction of future task requests can be realized. After the information is input into the low-altitude airspace resource configuration model, the unmanned aerial vehicle scheduling strategy can be optimized, the response speed and efficiency of resource configuration can be improved, and airspace conflicts and resource waste can be reduced.

[0030] The above description is only a summary of the technical solutions of the present application. In order to enable a more thorough understanding of the technical means of the present application, the present application can be implemented according to the content of the specification, and in order to enable the above and other purposes, characteristics and advantages of the present application to be more apparent and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0031] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of the preferred embodiments and are not meant to limit the present application. Moreover, the same reference numerals in the attached drawings indicate the same or similar components. In the drawings:

[0032] Figure 1 A flowchart of an embodiment of a low-altitude airspace resource configuration method provided by an embodiment of the present application is shown;

[0033] Figure 2 A flowchart of a first embodiment of a low-altitude airspace resource configuration method provided by an embodiment of the present application is shown;

[0034] Figure 3 A flowchart of a second embodiment of a low-altitude airspace resource configuration method provided by an embodiment of the present application is shown;

[0035] Figure 4 A structural diagram of a low-altitude airspace resource configuration device provided by an embodiment of the present application is shown;

[0036] Figure 5 A structural diagram of an electronic device provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0037] Exemplary embodiments of the present application will be described in greater detail below with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present application can be more thoroughly understood and the scope of the present application can be accurately conveyed to those skilled in the art.

[0038] With the rapid growth of the number of unmanned aerial vehicles, electric vertical takeoff and landing (eVTOL) aircraft and other aircraft, the development of low-altitude economy is facing unprecedented challenges, especially in terms of airspace management and resource allocation. The current management system mainly relies on manual division, static planning and centralized approval, which fails to adapt to the rapid changes and complexity of flight activities, resulting in significant technical defects.

[0039] Firstly, the resource configuration response lag is a prominent problem. Traditional scheduling methods are often based on fixed rules, which are difficult to quickly respond to dynamic factors such as weather changes, emergencies, and flight density. This lag not only leads to waste of resources, but also frequently causes air congestion, affecting overall flight safety and efficiency.

[0040] Secondly, the existing scheduling system lacks intelligent prediction and global optimization capabilities. Most airspace management systems are only responsible for task allocation and lack the ability to predict demand and model trends based on historical data, failing to form an effective technology loop. Such limitations prevent the system from anticipating potential resource conflicts and flight demand in advance, resulting in low scheduling efficiency.

[0041] In addition, the weak multi-target conflict handling capability is also a shortcoming of current technology. Existing methods often focus on a single target, but struggle to balance and coordinate between flight safety, airspace utilization, and task priority, making it difficult to achieve comprehensive optimization.

[0042] Finally, while some platforms attempt to introduce artificial intelligence algorithms, most still remain in the stage of auxiliary analysis and have not yet fully applied adaptive resource configuration and intelligent scheduling strategies based on deep learning and reinforcement learning. This situation limits the efficient configuration and safe scheduling capabilities of low-altitude airspace resources.

[0043] Based on this, the present application provides a low-altitude airspace resource configuration method, which integrates deep learning, reinforcement learning, and graph modeling artificial intelligence algorithms to achieve a closed-loop capability of prediction-scheduling-optimization, thereby realizing efficient configuration and safe scheduling of low-altitude airspace resources. The present application addresses scenarios with dense flights, complex airspace, and varying demands, enabling dynamic intelligent configuration of low-altitude airspace resources. Through deep learning and reinforcement learning algorithms, it can predict airspace usage trends based on real-time data and quickly reconstruct configuration schemes, significantly improving resource response speed and flexibility. At the same time, by mastering flight demand distribution through artificial intelligence prediction models and combining reinforcement learning strategies for scheduling, it improves airspace utilization and scheduling efficiency. Furthermore, the present application constructs a "prediction-configuration-evaluation-optimization" technology loop to support continuous policy evaluation and retraining, avoiding the risk of static planning failure. In addition, the present application can consider multiple dimensions in scheduling and automatically generate the optimal UAV scheduling strategy, ensuring efficient operation of UAV scheduling while ensuring safety and fairness.

[0044] As shown in Figure 1 The flowchart of one embodiment of the low-altitude airspace resource configuration method provided by the present application is shown in the figure. The method can include:

[0045] S100, obtaining real-time UAV operation data, real-time flight task queuing status and historical task flow data of the target city low-altitude airspace, wherein the real-time UAV operation data includes real-time UAV flight data, airspace map information, scheduling history data, current external variable data and real-time flight request data of the target city low-altitude airspace at the current time, and the historical task flow data includes historical external variable data and historical flight request data of the target city low-altitude airspace at at least one historical time.

[0046] Wherein, the target city low-altitude airspace refers to the low-altitude aviation space within a specific city range, used for the operation and flight of UAVs and other aircraft.

[0047] Wherein, the real-time UAV operation data refers to the flight status information of the UAV collected at the current time, including flight path, flight altitude, speed and heading, etc.

[0048] Wherein, the real-time flight task queuing status refers to the list of tasks to be executed by the UAV at the current time, including tasks waiting and their priority and status, reflecting the queuing situation of the tasks.

[0049] Wherein, the historical task flow data refers to the UAV flight task data recorded at at least one historical time, including past flight requests, task execution conditions and related parameters.

[0050] Wherein, the real-time UAV flight data includes GPS coordinates, altitude, speed and heading of the UAV, etc., used to monitor the real-time flight status of the UAV.

[0051] Wherein, the airspace map information refers to the geographical information about the flight area, indicating the flyable area, restricted area and dangerous area, etc., helping the UAV to avoid danger and comply with regulations during flight.

[0052] Wherein, the scheduling history data records the execution information of past tasks of the UAV, including execution time, flight path and task completion rate of the task, used to analyze and optimize future scheduling strategies.

[0053] Wherein, the current external variable data is the real-time external factor affecting the flight of the UAV, including weather, wind speed and visibility, etc.

[0054] Wherein, the real-time flight request data refers to the UAV flight task request currently submitted by the user, including the starting point, end point, purpose, time requirement, etc.

[0055] Wherein, the historical external variable data is the external variable data collected at a certain time in the past, including historical weather conditions, wind speed and visibility, etc., used to analyze the trend of changes in flight conditions.

[0056] The historical flight request data refers to a record of a UAV flight task request submitted by a user at a historical time, including detailed information of the task and an execution result, and is used for evaluating and predicting subsequent task requirements.

[0057] Specifically, the embodiment of the present application can interface with the ADS-B (Automatic Dependent Surveillance-Broadcast) of the UAV, collect GPS coordinates, height, speed and heading information of each UAV in real time, to reflect its current flight state and monitor the position and dynamics. At the same time, access to the city airspace map, including information of flyable area, restricted area and danger area, to judge the legality and safety of UAV flight. In addition, the scheduling history data records the execution of past tasks of the UAV, such as task execution time, flight path and completion rate, so as to optimize future task scheduling. At the same time, real-time monitoring of external variable data, such as weather, wind speed and visibility, to ensure the safety and efficiency of UAV operation, especially in special weather conditions. Finally, receiving real-time flight requests submitted by users, including take-off and landing points, time window and load, to update the queuing status and scheduling plan of the task.

[0058] S110, obtaining a current low-altitude airspace resource state graph of the target city low-altitude airspace by using real-time UAV operation data, wherein the current low-altitude airspace resource state graph includes airspace feature vectors of a plurality of airspace units in an airspace adjacency graph constructed based on the target city low-altitude airspace.

[0059] The current low-altitude airspace resource state graph is a comprehensive graphical representation reflecting the resource state of each airspace unit in the target city low-altitude airspace. The current low-altitude airspace resource state graph is generated by real-time UAV operation data, contains airspace feature vectors of a plurality of airspace units, and provides information about aircraft distribution, flight density, historical congestion, weather factors and airspace importance, providing decision support for UAV scheduling and management.

[0060] The airspace adjacency graph is obtained by modeling the airspace state of the target city low-altitude airspace based on a graph neural network. The nodes in the airspace adjacency graph represent airspace units, and the edges in the airspace adjacency graph represent physical or scheduling connections between adjacent airspace units. Specifically, the airspace adjacency graph is a graph structure composed of a plurality of airspace units. Each node in the graph represents an airspace unit, and the edges between the nodes represent the physical or scheduling connection relationship between adjacent airspace units. By constructing the airspace adjacency graph structure, the mutual relationship between the airspace units can be effectively described and analyzed, providing a basis for graph neural network modeling.

[0061] The airspace unit refers to an independent dispatch unit formed after the city low-altitude airspace is rasterized, and is usually divided in the form of a 100m*100m grid. Each airspace unit contains information about the number of drones in the region, flight density, historical congestion coefficient, weather factor, and airspace importance, for real-time monitoring and scheduling decisions.

[0062] The airspace feature vector is the result of vectorizing each airspace unit, containing features such as current aircraft number, flight density, historical congestion coefficient, weather factor, and airspace importance coefficient. The airspace feature vector is extracted through a graph neural network.

[0063] Specifically, the embodiments of the present application can divide the target city low-altitude airspace into 100m*100m grid units, and each grid unit can be regarded as an independent dispatch unit. This rasterization makes the management and scheduling of airspace more detailed and accurate, and helps to monitor the flight status and distribution of drones in real time. After completing the airspace rasterization, the next step is to build an airspace adjacency graph structure. This structure regards each airspace unit as a node in the graph, and adjacent airspace units are connected by edges. This adjacency graph can effectively describe the physical or scheduling relationship between airspace units, providing a basis for subsequent data analysis and modeling. Each airspace unit is represented in the form of a vector, containing multiple important indicators, including: current aircraft number, flight density, historical congestion coefficient, weather factor, and airspace importance coefficient.

[0064] The current aircraft number refers to the number of drones in the airspace unit.

[0065] The flight density represents the number of flight tasks per unit area, reflecting the use of airspace.

[0066] The historical congestion coefficient is used to evaluate the congestion level of the airspace.

[0067] The weather factor includes wind speed, rainfall, and other meteorological conditions that affect the flight safety of drones.

[0068] The airspace importance coefficient is used to consider the influence of factors such as nearby airports and emergency routes, and to evaluate the importance of the airspace unit.

[0069] Specifically, the embodiment of the present application can utilize a graph neural network (GNN) to embed and encode the constructed airspace adjacency graph. The graph neural network can effectively capture the structural information and feature relationships between airspace units, where the graph nodes represent airspace units, and each node carries its corresponding state vector. The edges represent the physical or scheduling connections between adjacent units. The input features include task popularity, available bandwidth, security level, etc., which help better analyze and understand the airspace state. Finally, the real-time state vector of each airspace unit is extracted, and the current low-altitude airspace resource state graph is formed. The current low-altitude airspace resource state graph contains the airspace feature vectors of multiple airspace units, comprehensively reflecting the resource state of the urban low-altitude airspace, and providing an important basis for the scheduling, management and decision-making of unmanned aerial vehicles.

[0070] S120, input the historical task flow data into the pre-constructed flight demand prediction model to obtain a flight task demand prediction heat map of the target urban low-altitude airspace output by the flight demand prediction model, wherein the flight task demand prediction heat map includes the task request prediction results of each airspace unit in a future period of time.

[0071] The flight demand prediction model is constructed based on a deep learning neural network and improves the prediction spatial correlation through a spatio-temporal graph convolution network. Specifically, the flight demand prediction model is a kind of model that utilizes machine learning and deep learning techniques, especially time series analysis methods (such as LSTM (Long Short-Term Memory Network) or Transformer), to model historical flight request data and related meteorological information, in order to predict the flight task demand of each airspace unit in a future period of time. The flight demand prediction model can generate predictions of future demand by analyzing patterns and trends in historical data, thereby providing a scientific basis for unmanned aerial vehicle scheduling and management.

[0072] The flight task demand prediction heat map refers to a visual image generated by the flight demand prediction model, which shows the task request prediction results of each airspace unit in the target urban low-altitude airspace in a future period of time. The heat map usually represents the task request density or heat of different airspace units with color depth or lightness, which can intuitively show which areas are expected to have higher flight task demand, thereby helping managers identify potential congestion areas.

[0073] The task request prediction result is the specific data output by the flight demand prediction model, representing the predicted number of flight task requests in each airspace unit at a specific future time. This result is usually presented in numerical form, reflecting the number of flight tasks each airspace unit is expected to receive within a certain future time. Through these predictions, the scheduling system can better allocate resources and adjust to potential flight task peaks.

[0074] To improve the understanding of spatial correlation by the flight demand prediction model, a spatio-temporal graph convolution network is used to process both time and spatial information, enhancing the prediction ability of flight demand in the city. The spatio-temporal graph convolution network can capture the common "hot spot migration effect" in the city, such as during peak hours, it is still possible to observe the phenomenon of flight demand transferring from busy areas to surrounding areas.

[0075] After the training of the flight demand prediction model is completed, historical task flow data is input into the flight demand prediction model, which will output the predicted number of task requests for each airspace unit within a certain future time (e.g., the next 30 minutes), generating a prediction value for each airspace unit. These prediction values are aggregated to form a flight task demand prediction heat map. The flight task demand prediction heat map visually displays the flight task demand of each airspace unit in the future, usually using color depth or lightness to represent the density of requests, thus not only helping to optimize the allocation of airspace resources, but also identifying potential congestion areas in advance, so that preventive measures can be taken to ensure flight safety and efficiency.

[0076] S130, input the real-time flight task queuing state, the current low-altitude airspace resource state map, and the flight task demand prediction heat map into the pre-constructed low-altitude airspace resource configuration model to obtain the unmanned aerial vehicle scheduling strategy output by the low-altitude airspace resource configuration model.

[0077] The low-altitude airspace resource configuration model is a decision-making model based on real-time data and prediction information, used to optimize the efficiency of unmanned aerial vehicles operating in low-altitude airspace.

[0078] Optionally, the low-altitude airspace resource configuration model is constructed using a proximal policy optimization algorithm, and a reward function R = λ1U - λ2C - λ3D is used to generate a UAV scheduling strategy, where U is resource utilization efficiency, C is the number of airspace conflicts, D is the average task delay time, and λ1, λ2 and λ3 are adjustable weight coefficients. The low-altitude airspace resource configuration model integrates the current flight task queuing state, low-altitude airspace resource state diagram and flight task demand prediction heat map, and uses a proximal policy optimization (PPO) algorithm to dynamically generate a UAV scheduling strategy. The goal of the low-altitude airspace resource configuration model is to maximize resource utilization efficiency, reduce airspace conflicts and task delays, and thus improve the safety and efficiency of overall flight management. In the training process of the low-altitude airspace resource configuration model, the Actor-Critic structure can be used for policy learning. Actor-Critic is a reinforcement learning framework that combines policy and value functions. It includes two main components: Actor (policy network) and Critic (value function network). The Actor is responsible for selecting actions (scheduling strategies), while the Critic evaluates the value of the action (i.e. expected reward). Through the Actor-Critic structure, the low-altitude airspace resource configuration model can more efficiently learn and optimize the scheduling strategy.

[0079] Optionally, the embodiments of the present application can use a rule engine (such as Drools) to define initial airspace resource scheduling rules. The rule engine can dynamically manage UAV tasks according to specific conditions and constraints (such as flight safety, flight altitude restrictions, airspace usage priority, etc.). The completion rate and conflict risk of flight tasks are predicted using traditional machine learning algorithms (such as XGBoost and Random Forest). In each round of UAV scheduling, the scheduling rules defined in the rule base of the rule engine are combined with the UAV scheduling strategy predicted by the low-altitude airspace resource configuration model to dynamically adjust real-time UAV scheduling, so that the UAV scheduling can adapt to real-time changes in the environment and demand.

[0080] The UAV scheduling strategy refers to the specific action plan and instructions output by the low-altitude airspace resource configuration model, used to guide the behavior adjustment of the UAV when executing flight tasks. The UAV scheduling strategy can include congestion area fly-around suggestions, take-off delay or early instructions, and path redirection and time window reordering.

[0081] Specifically, in this embodiment of the invention, after inputting the real-time flight mission queuing status, the current low-altitude airspace resource status map, and the flight mission demand prediction heatmap into a pre-constructed low-altitude airspace resource allocation model, the low-altitude airspace resource allocation model first performs state encoding on the input data, transforming the real-time flight mission queuing status, the current low-altitude airspace resource status map, and the flight mission demand prediction heatmap into a form that can be processed by the algorithm. Then, using an Actor-Critic structure, the model combines the state input with the current policy network to generate corresponding actions. These actions can include priority ranking of airspace units, airspace resource reallocation strategies, and conflict path avoidance strategies. After processing, the low-altitude airspace resource allocation model outputs UAV scheduling strategies including congestion area detour suggestions, takeoff delay or early takeoff instructions, and path redirection and time window reordering.

[0082] This invention provides a method for allocating low-altitude airspace resources. The method includes: obtaining real-time UAV operation data, real-time flight mission queuing status, and historical mission flow data for the low-altitude airspace of a target city. The real-time UAV operation data includes real-time UAV flight data, airspace map information, scheduling history data, current external variable data, and real-time flight request data for the target city's low-altitude airspace at the current moment. The historical mission flow data includes historical external variable data and historical flight request data for the target city's low-altitude airspace at at least one historical moment. Using the real-time UAV operation data, a current low-altitude airspace resource status map of the target city's low-altitude airspace is obtained. The airspace resource status map includes airspace feature vectors of multiple airspace units in an airspace adjacency graph constructed based on the low-altitude airspace of the target city. Historical task flow data is input into a pre-built flight demand prediction model to obtain a flight task demand prediction heatmap of the target city's low-altitude airspace, output by the model. This heatmap includes the predicted task requests for each airspace unit over a future period. Real-time flight task queuing status, the current low-altitude airspace resource status map, and the flight task demand prediction heatmap are input into a pre-built low-altitude airspace resource allocation model to obtain the UAV scheduling strategy output by the model. This invention, by collecting UAV operation data, flight task queuing status, and historical task flow data in real time, establishes the current low-altitude airspace resource status map and the flight task demand prediction heatmap, enabling accurate prediction of future task requests. Inputting this information into the low-altitude airspace resource allocation model optimizes UAV scheduling strategies, improves the response speed and efficiency of resource allocation, and reduces airspace conflicts and resource waste.

[0083] Optional, based on Figure 1 The method shown is as follows: Figure 2As shown in the flowchart of the first embodiment of the low-altitude airspace resource configuration method provided by the embodiments of the present application, after step S130, the method can further include:

[0084] S200, dispatching the UAV scheduling strategy to the UAV operation platform, so that the UAV operation platform performs the UAV flight task in the target city low-altitude airspace according to the UAV scheduling strategy.

[0085] Specifically, the embodiments of the present application can be connected with the UAV operation platform through the control interface, and the UAV scheduling strategy is transmitted to the UAV operation platform by the control interface. The UAV scheduling strategy is encoded in a specific format that can be understood and processed by the UAV operation platform, and is transmitted to the UAV operation platform after security verification. After the UAV operation platform analyzes the received UAV scheduling strategy, it will identify the specific flight task, target path and any restrictions that need to be followed, and will assign the flight task to the corresponding UAV. Each UAV will receive its specific flight task, takeoff time, path and other execution instructions. The UAV performs the UAV flight task in the target city low-altitude airspace according to the UAV scheduling strategy. The UAV flight task can specifically include taking off as planned, flying to a designated location, performing a task (such as monitoring, data collection, transportation, etc.), and returning on time or completing the task.

[0086] It can be understood that during the execution of the task by the UAV, the UAV operation platform will usually perform real-time monitoring to ensure that the UAV flies safely according to the scheduling strategy. If abnormal situations (such as weather changes, unexpected tasks, etc.) occur during execution, the current situation will be re-evaluated and the scheduling strategy will be dynamically adjusted.

[0087] The embodiments of the present application not only ensure that the UAV can perform the task, but also enhance the flexibility of airspace management and the ability to respond to unexpected situations, thereby improving the overall safety and efficiency of the UAV in performing tasks in the city low-altitude airspace.

[0088] Optionally, based on Figure 2 As shown in the method, as Figure 3 As shown in the flowchart of the second embodiment of the low-altitude airspace resource configuration method provided by the embodiments of the present application, after step S200, the method can further include:

[0089] S300, obtaining actual scheduling execution data of the UAV flight task, wherein the actual scheduling execution data includes task completion status, flight conflict rate and delay quantity of the UAV flight task.

[0090] The actual scheduling execution data refers to the execution record of the UAV flight task collected during actual operation.

[0091] The task completion condition refers to the number of tasks successfully performed and completed by the unmanned aerial vehicle within a specified time, and is used to reflect the effectiveness and execution efficiency of the unmanned aerial vehicle scheduling strategy.

[0092] The flight conflict rate refers to the frequency of flight conflict events occurring during the flight of the unmanned aerial vehicle, that is, how many times the unmanned aerial vehicle has path overlap or potential conflict caused by proximity in the airspace, and is used to determine the airspace scheduling safety of the unmanned aerial vehicle scheduling strategy.

[0093] The number of delays refers to the number of times the unmanned aerial vehicle fails to take off or complete a task as planned due to various reasons such as airspace congestion, equipment failure, weather changes, etc., and is used to determine the user experience and task completion rate of the unmanned aerial vehicle scheduling strategy.

[0094] Specifically, the embodiments of the present application can monitor the flight state of the unmanned aerial vehicle in real time through the unmanned aerial vehicle operation platform, record the execution of each task, such as the timestamps of task start and completion, whether it is completed on time, the flight conflict situation encountered (for example, the number of times the flight path overlaps), and the number of task delays caused by various factors (such as weather, airspace restrictions, etc.). The embodiments of the present application can use sensor data, flight logs and feedback from the scheduling system to aggregate actual scheduling execution data, realizing real-time collection and storage of data.

[0095] S310, based on the actual scheduling execution data, calculating evaluation index detection data of the unmanned aerial vehicle scheduling strategy, wherein the evaluation index detection data includes airspace utilization rate, user satisfaction score and system stability.

[0096] The evaluation index detection data refers to important indicators obtained after analyzing the actual scheduling execution data, and is used to evaluate the effect of the unmanned aerial vehicle scheduling strategy.

[0097] The airspace utilization rate refers to the ratio of the actual executed unmanned aerial vehicle task density to the maximum carrying density of the airspace within a certain time, and is used to determine whether the current unmanned aerial vehicle scheduling strategy effectively utilizes low-altitude airspace resources.

[0098] The user satisfaction score refers to a score obtained based on user feedback on unmanned aerial vehicle scheduling services, taking into account factors such as scheduling response time and task completion rate, and is used to reflect the user's satisfaction with the quality of scheduling services under the current unmanned aerial vehicle scheduling strategy.

[0099] The system stability refers to whether the current unmanned aerial vehicle scheduling frequently switches strategies during task execution, and whether it causes re-scheduling due to flight conflicts. High stability indicates that the unmanned aerial vehicle scheduling system runs more smoothly and has a low conflict probability.

[0100] Specifically, the embodiment of the present application can calculate the evaluation index detection data through data analysis tools and algorithms. Using the task completion, the success rate of the task and the airspace utilization rate (the ratio of the number of tasks completed to the total number of planned tasks) can be calculated. The user satisfaction score can be obtained through a questionnaire survey or a user feedback system, and factors such as scheduling response time and task completion rate are analyzed. The system stability is evaluated by counting the flight conflict rate and the strategy change frequency.

[0101] S320, updating the reward function of the low-altitude airspace resource configuration model based on the evaluation index detection data to obtain an updated low-altitude airspace resource configuration model.

[0102] The embodiment of the present application can convert the values of each index in the evaluation index detection data into a reward signal. For example, the improvement of airspace utilization rate and user satisfaction can correspond to a positive reward, while the increase of flight conflict rate and delay quantity corresponds to a negative reward. The reward function of the model is optimized using the reinforcement learning algorithm, so that it is more effective and intelligent in processing future low-altitude airspace resource scheduling, thereby helping the low-altitude airspace resource configuration model to adapt to historical data and continuously improve the quality of generated unmanned aerial vehicle scheduling strategies.

[0103] The evaluation index detection data calculated based on the actual scheduling execution data can provide a quantitative basis for the evaluation of unmanned aerial vehicle scheduling strategies, and the evaluation index detection data is used to update the reward function of the low-altitude airspace resource configuration model, which can enable the low-altitude airspace resource configuration model to continuously learn and optimize in actual operation, better cope with future flight demand and airspace changes, and quickly adjust the strategy to reduce the flight conflict rate and improve the task completion rate in the face of unexpected situations, thereby ensuring the safety and efficiency of unmanned aerial vehicle operation.

[0104] Optionally, in the above Figure 1 Based on one or more corresponding embodiments, another optional embodiment provided by the embodiment of the present application can further include, before step S110, the method can further include:

[0105] The real-time unmanned aerial vehicle operation data is subjected to missing value filling and data standardization processing.

[0106] The missing value filling refers to supplementing the data points with blanks or missing values in the real-time unmanned aerial vehicle operation data through certain methods to ensure the integrity and continuity of the data.

[0107] Optionally, the embodiment of the present application can supplement the data points with blanks or missing values in the real-time unmanned aerial vehicle operation data based on the nearest neighbor algorithm. For example, for a missing unmanned aerial vehicle flight speed, the flight speeds in the previous and subsequent time periods can be found, and the average of these nearest values can be used to fill in the missing values.

[0108] Optionally, the embodiment of the present application can supplement the missing data points in the real-time unmanned aerial vehicle operation data based on a linear interpolation algorithm. For example, if the height of the unmanned aerial vehicle at a certain time is missing, the missing value can be calculated according to the height values before and after the time.

[0109] The data standardization refers to unifying the data of different features in the real-time unmanned aerial vehicle operation data on the same scale for comparison and analysis. Optionally, the data standardization algorithm can include a Z-score standardization algorithm and a Min-Max normalization algorithm.

[0110] The embodiment of the present application fills in the missing values and performs data standardization processing on the real-time unmanned aerial vehicle operation data, ensures high-quality data input, and enables the airspace resource state graph generated to accurately reflect the current airspace conditions, thereby providing a more reliable basis for subsequent flight demand prediction and unmanned aerial vehicle scheduling strategy formulation.

[0111] Optionally, the embodiment of the present application can train and update the model in the cloud computing platform, and deploy the latest model in the edge computing node (such as the airport ground dispatch station), and through API (Application Programming Interface) connection with the unmanned traffic management system (UTM) or the unmanned traffic dispatch platform, to realize real-time interaction and scheduling optimization of data.

[0112] Although the operations are depicted in a particular order, this should not be understood as requiring the operations to be performed in the particular order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing can be advantageous.

[0113] It should be understood that each of the steps described in the method embodiments of the present application can be performed in different orders and / or in parallel. In addition, the method embodiments can include additional steps and / or omit the execution of the steps shown. The scope of the present application is not limited in this respect.

[0114] Corresponding to the above-mentioned method embodiments, the embodiment of the present application also provides a low-altitude airspace resource configuration device, the structure of which is as shown in Figure 4 The device can include a data acquisition unit 10, a current low-altitude airspace resource state graph obtaining unit 20, a flight task demand prediction heat map obtaining unit 30, and an unmanned aerial vehicle scheduling strategy obtaining unit 40.

[0115] The data acquisition unit 10 is configured to obtain real-time unmanned aerial vehicle operation data, real-time flight task queuing state and historical task flow data of the target city low-altitude airspace, wherein the real-time unmanned aerial vehicle operation data comprises real-time unmanned aerial vehicle flight data, airspace map information, scheduling historical data, current external variable data and real-time flight request data of the target city low-altitude airspace at a current time, and the historical task flow data comprises historical external variable data and historical flight request data of the target city low-altitude airspace at at least one historical time.

[0116] The current low-altitude airspace resource state diagram acquisition unit 20 is configured to obtain a current low-altitude airspace resource state diagram of the target city low-altitude airspace by using the real-time unmanned aerial vehicle operation data, wherein the current low-altitude airspace resource state diagram comprises airspace feature vectors of a plurality of airspace units in an airspace adjacency graph constructed based on the target city low-altitude airspace.

[0117] The flight task demand prediction heat map acquisition unit 30 is configured to input the historical task flow data into a pre-constructed flight demand prediction model to obtain a flight task demand prediction heat map of the target city low-altitude airspace output by the flight demand prediction model, wherein the flight task demand prediction heat map comprises a task request prediction result of each airspace unit in a future period of time.

[0118] The unmanned aerial vehicle scheduling strategy acquisition unit 40 is configured to input the real-time flight task queuing state, the current low-altitude airspace resource state diagram and the flight task demand prediction heat map into a pre-constructed low-altitude airspace resource configuration model to obtain a unmanned aerial vehicle scheduling strategy output by the low-altitude airspace resource configuration model.

[0119] Optionally, the low-altitude airspace resource configuration apparatus can further comprise a unmanned aerial vehicle scheduling strategy issuing unit.

[0120] The unmanned aerial vehicle scheduling strategy issuing unit is configured to, after the unmanned aerial vehicle scheduling strategy acquisition unit 40 inputs the real-time flight task queuing state, the current low-altitude airspace resource state diagram and the flight task demand prediction heat map into the pre-constructed low-altitude airspace resource configuration model to obtain the unmanned aerial vehicle scheduling strategy output by the low-altitude airspace resource configuration model, issue the unmanned aerial vehicle scheduling strategy to a unmanned aerial vehicle operation platform, so that the unmanned aerial vehicle operation platform performs unmanned aerial vehicle flight tasks in the target city low-altitude airspace according to the unmanned aerial vehicle scheduling strategy.

[0121] Optionally, the low-altitude airspace resource configuration apparatus can further comprise an actual scheduling execution data acquisition unit, an evaluation index detection data acquisition unit and a low-altitude airspace resource configuration model updating unit.

[0122] The actual scheduling execution data obtaining unit is configured to obtain actual scheduling execution data of the UAV flight task after the UAV operation platform executes the UAV flight task in the target city low-altitude airspace according to the UAV scheduling strategy, wherein the actual scheduling execution data comprises task completion, flight conflict rate and delay quantity of the UAV flight task.

[0123] The evaluation index detection data obtaining unit is configured to calculate evaluation index detection data of the UAV scheduling strategy based on the actual scheduling execution data, wherein the evaluation index detection data comprises airspace utilization rate, user satisfaction score and system stability.

[0124] The low-altitude airspace resource configuration model updating unit is configured to update the reward function of the low-altitude airspace resource configuration model based on the evaluation index detection data, and obtain an updated low-altitude airspace resource configuration model.

[0125] Optionally, the low-altitude airspace resource configuration device can further comprise a real-time UAV operation data processing unit.

[0126] The real-time UAV operation data processing unit is configured to perform missing value filling and data standardization processing on real-time UAV operation data before the current low-altitude airspace resource state map obtaining unit 20 obtains the current low-altitude airspace resource state map of the target city low-altitude airspace by using the real-time UAV operation data.

[0127] Optionally, the airspace adjacency graph is obtained by modeling the airspace state of the target city low-altitude airspace based on a graph neural network, wherein a node in the airspace adjacency graph represents an airspace unit, and an edge in the airspace adjacency graph represents physical or scheduling connection between two adjacent airspace units.

[0128] Optionally, the flight demand prediction model is constructed based on a deep learning neural network, and the spatial correlation is improved by using a spatio-temporal graph convolution network.

[0129] Optionally, the low-altitude airspace resource configuration model is constructed by using a proximal policy optimization algorithm, and a UAV scheduling strategy is generated based on a reward function R = λ1U - λ2C - λ3D, wherein U is resource utilization efficiency, C is airspace conflict quantity, D is average task delay time, and λ1, λ2 and λ3 are adjustable weight coefficients.

[0130] The low-altitude airspace resource configuration device provided by the application is used to obtain real-time unmanned aerial vehicle operation data, real-time flight task queuing state and historical task flow data of a target city low-altitude airspace, wherein the real-time unmanned aerial vehicle operation data comprises real-time unmanned aerial vehicle flight data, airspace map information, scheduling historical data, current external variable data and real-time flight request data of the target city low-altitude airspace at a current time, and the historical task flow data comprises historical external variable data and historical flight request data of the target city low-altitude airspace at at least one historical time; the current low-altitude airspace resource state graph of the target city low-altitude airspace is obtained by using the real-time unmanned aerial vehicle operation data, wherein the current low-altitude airspace resource state graph comprises airspace feature vectors of a plurality of airspace units in an airspace adjacency graph constructed based on the target city low-altitude airspace; the historical task flow data is input into a flight demand prediction model constructed in advance to obtain a flight task demand prediction heat map of the target city low-altitude airspace output by the flight demand prediction model, wherein the flight task demand prediction heat map comprises a task request prediction result of each airspace unit in a future period of time; the real-time flight task queuing state, the current low-altitude airspace resource state graph and the flight task demand prediction heat map are input into a low-altitude airspace resource configuration model constructed in advance to obtain an unmanned aerial vehicle scheduling strategy output by the low-altitude airspace resource configuration model. The application can realize accurate prediction of future task requests by collecting real-time unmanned aerial vehicle operation data, flight task queuing state and historical task flow data, establishing a current low-altitude airspace resource state graph and a flight task demand prediction heat map. After inputting the information into the low-altitude airspace resource configuration model, the unmanned aerial vehicle scheduling strategy can be optimized, the response speed and efficiency of resource configuration can be improved, and airspace conflicts and resource waste can be reduced.

[0131] As to the device in the above-mentioned embodiments, the specific manner in which each unit performs operations has been described in detail in the embodiments related to the method, and thus will not be described in detail here.

[0132] The low-altitude airspace resource configuration device comprises a processor and a memory, and the above-mentioned data acquisition unit 10, current low-altitude airspace resource state graph obtaining unit 20, flight task demand prediction heat map obtaining unit 30 and unmanned aerial vehicle scheduling strategy obtaining unit 40 are stored in the memory as program units, and the corresponding functions are realized by the processor executing the above-mentioned program units stored in the memory.

[0133] The processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and by adjusting kernel parameters, real-time data on UAV operations, flight mission queuing status, and historical mission flow can be collected. This allows for the creation of a current low-altitude airspace resource status map and a flight mission demand prediction heatmap, enabling accurate prediction of future mission requests. Inputting this information into the low-altitude airspace resource allocation model optimizes UAV scheduling strategies, improves the response speed and efficiency of resource allocation, and reduces airspace conflicts and resource waste.

[0134] This invention provides a computer-readable storage medium storing a program that, when executed by a processor, implements the low-altitude airspace resource allocation method.

[0135] This invention provides a processor for running a program, wherein the program executes the low-altitude airspace resource configuration method during runtime.

[0136] like Figure 5 As shown, this embodiment of the invention provides an electronic device 1000, which includes at least one processor 1001, at least one memory 1002 connected to the processor 1001, and a bus 1003. The processor 1001 and the memory 1002 communicate with each other via the bus 1003. The processor 1001 is used to call program instructions in the memory 1002 to execute the aforementioned low-altitude airspace resource allocation method. The electronic device in this document can be a server, PC, PAD, mobile phone, etc.

[0137] The present invention also provides a computer program product that, when executed on an electronic device, is suitable for executing a program that initializes a method for configuring low-altitude airspace resources.

[0138] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatuses, electronic devices (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0139] In a typical configuration, an electronic device includes one or more processors (CPUs), memory, and a bus. The electronic device may also include input / output interfaces, network interfaces, etc.

[0140] Memory can include non-persistent memory, Random Access Memory (RAM), and / or non-volatile memory such as flash memory, Read Only Memory (ROM), or Electrically Programmable Read Only Memory (EPROM), etc., in a computer readable medium. Memory includes at least one memory chip. Memory is an example of computer readable media.

[0141] Computer readable media includes permanent and non-permanent, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory (ROM), electrically programmable read only memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital versatile discs (DVDs) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible to a computing device. According to the definition herein, computer readable media does not include transitory media such as modulated data signals and carrier waves.

[0142] In the description of the present application, it needs to be understood that if the orientation or positional relationship indicated by the terms such as "upper", "lower", "front", "back", "left" and "right" is based on the orientation or positional relationship shown in the drawings, it is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the indicated position or element must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application.

[0143] It should be noted that in this paper, the relationship terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations. It should also be noted that the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, product or equipment including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, product or equipment. Without more limitation, the element defined by the statement "including a" does not exclude the presence of another identical element in the process, method, product or equipment including the element.

[0144] Those skilled in the art will appreciate that embodiments of the present application can be devised for a method, a system, or a computer program product. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer-readable program code thereon for use by or in connection with an instruction execution system. For the purposes of this description, a computer-usable or computer readable storage medium can be any apparatus that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.

[0145] The above merely provides an embodiment of the present application and is not intended to limit the present application. The present application can have various modifications and changes for those skilled in the art. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the scope of the present application.

Claims

1. A method for low-altitude airspace resource configuration, characterized in that, The method comprises the following steps: obtaining real-time unmanned aerial vehicle operation data, real-time flight task queuing state and historical task flow data of a target city low-altitude airspace, wherein the real-time unmanned aerial vehicle operation data comprises real-time unmanned aerial vehicle flight data, airspace map information, scheduling historical data, current external variable data and real-time flight request data of the target city low-altitude airspace at a current time, and the historical task flow data comprises historical external variable data and historical flight request data of the target city low-altitude airspace at at least one historical time; obtaining a current low-altitude airspace resource state graph of the target city low-altitude airspace by using the real-time unmanned aerial vehicle operation data, wherein the current low-altitude airspace resource state graph comprises airspace feature vectors of a plurality of airspace units in an airspace adjacency graph constructed based on the target city low-altitude airspace; inputting the historical task flow data into a pre-constructed flight demand prediction model to obtain a flight task demand prediction heat map of the target city low-altitude airspace output by the flight demand prediction model, wherein the flight task demand prediction heat map comprises a task request prediction result of each airspace unit within a future period of time; inputting the real-time flight task queuing state, the current low-altitude airspace resource state graph and the flight task demand prediction heat map into a pre-constructed low-altitude airspace resource configuration model to obtain an unmanned aerial vehicle scheduling strategy output by the low-altitude airspace resource configuration model.

2. The method of claim 1, wherein, After the real-time flight task queuing state, the current low-altitude airspace resource state graph and the flight task demand prediction heat map are inputted into the pre-constructed low-altitude airspace resource configuration model to obtain the unmanned aerial vehicle scheduling strategy output by the low-altitude airspace resource configuration model, the method further comprises: issuing the unmanned aerial vehicle scheduling strategy to an unmanned aerial vehicle operation platform, so that the unmanned aerial vehicle operation platform executes unmanned aerial vehicle flight tasks in the target city low-altitude airspace according to the unmanned aerial vehicle scheduling strategy.

3. The method of claim 2, wherein, After the unmanned aerial vehicle scheduling strategy is issued to the unmanned aerial vehicle operation platform, so that the unmanned aerial vehicle operation platform executes unmanned aerial vehicle flight tasks in the target city low-altitude airspace according to the unmanned aerial vehicle scheduling strategy, the method further comprises: obtaining actual scheduling execution data of the unmanned aerial vehicle flight tasks, wherein the actual scheduling execution data comprises task completion, flight conflict rate and delay quantity of the unmanned aerial vehicle flight tasks; calculating evaluation index detection data of the unmanned aerial vehicle scheduling strategy based on the actual scheduling execution data, wherein the evaluation index detection data comprises airspace utilization rate, user satisfaction score and system stability; updating a reward function of the low-altitude airspace resource configuration model based on the evaluation index detection data to obtain an updated low-altitude airspace resource configuration model.

4. The method of claim 1, wherein, Before the current low-altitude airspace resource state graph of the target city low-altitude airspace is obtained by using the real-time unmanned aerial vehicle operation data, the method further comprises: performing missing value filling and data standardization processing on the real-time unmanned aerial vehicle operation data.

5. The method of claim 1, wherein, The airspace adjacency graph is obtained by modeling airspace states of the target city low-altitude airspace based on a graph neural network, nodes in the airspace adjacency graph represent the airspace units, and edges in the airspace adjacency graph represent physical or scheduling connections between two adjacent airspace units.

6. The method of claim 1, wherein, The flight demand prediction model is constructed based on a deep learning neural network and improves prediction spatial correlation through a spatio-temporal graph convolution network.

7. The method of claim 1, wherein, The low-altitude airspace resource configuration model is constructed using a proximal policy optimization algorithm, and generates the unmanned aerial vehicle scheduling strategy based on a reward function R = λ 1 U - λ 2 C - λ 3 D, wherein U is resource utilization efficiency, C is airspace conflict quantity, D is average task delay time, and λ 1, λ 2 and λ 3 are adjustable weight coefficients.

8. A low-altitude airspace resource configuration apparatus, characterized by, Comprise: a data acquisition unit, a current low-altitude airspace resource state graph obtaining unit, a flight task demand prediction heat map obtaining unit, and an unmanned aerial vehicle scheduling strategy obtaining unit, The data acquisition unit is configured to obtain real-time unmanned aerial vehicle operation data, real-time flight task queuing state, and historical task flow data of the target city low-altitude airspace, wherein the real-time unmanned aerial vehicle operation data comprises real-time unmanned aerial vehicle flight data, airspace map information, scheduling history data, current external variable data, and real-time flight request data of the target city low-altitude airspace at a current time, and the historical task flow data comprises historical external variable data and historical flight request data of the target city low-altitude airspace at at least one historical time; The current low-altitude airspace resource state graph obtaining unit is configured to obtain a current low-altitude airspace resource state graph of the target city low-altitude airspace by using the real-time unmanned aerial vehicle operation data, wherein the current low-altitude airspace resource state graph comprises airspace feature vectors of a plurality of airspace units in an airspace adjacency graph constructed based on the target city low-altitude airspace; The flight task demand prediction heat map obtaining unit is configured to input the historical task flow data into a pre-constructed flight demand prediction model, and obtain a flight task demand prediction heat map of the target city low-altitude airspace output by the flight demand prediction model, wherein the flight task demand prediction heat map comprises task request prediction results of each airspace unit within a future period of time; The unmanned aerial vehicle scheduling strategy obtaining unit is configured to input the real-time flight task queuing state, the current low-altitude airspace resource state graph, and the flight task demand prediction heat map into a pre-constructed low-altitude airspace resource configuration model, and obtain an unmanned aerial vehicle scheduling strategy output by the low-altitude airspace resource configuration model.

9. A computer-readable storage medium having stored thereon a program, characterized in that, The program is executed by a processor to implement the low-altitude airspace resource configuration method of any one of claims 1 to 7.

10. An electronic device, comprising: The electronic device comprises at least one processor, at least one memory connected to the processor, and a bus; wherein the processor, the memory complete mutual communication through the bus; the processor is used to call the program instruction in the memory, in order to execute the low-altitude airspace resource configuration method of any one of claims 1 to 7.

Images