Low-altitude airspace channel setting method for urban air traffic scene

Abstract

The invention discloses a low-altitude airspace channel setting method for an urban air traffic scene, and relates to the technical field of airspace channels, and the method comprises the steps: collecting the operation data of an aircraft in real time, subdividing the flight height layer of the aircraft, and preliminarily dividing the urban airspace; the method comprises the following steps: forming aircraft trajectory sequence data based on trajectory data, performing trajectory sequence point mapping, calculating the similarity of aircraft trajectories by using a DTW algorithm, and performing clustering analysis on DTW similarity data by using a density peak clustering DPC algorithm. According to the method, density analysis is carried out on a DTW similarity matrix through DPC clustering, a density peak value in a track is found to serve as a clustering center, a highly flexible clustering mechanism is formed, efficient division of the airspace is achieved through further airspace division, and the clustering accuracy is improved based on a clustering analysis result and input state space data. The DDPG model can dynamically predict the maximum number of aircrafts which can be accommodated in the current airspace and optimize the real-time path planning of the aircrafts.

Description

Technical Field

[0001] The present invention relates to the field of airspace channel technology, and in particular to a low-altitude airspace channel demarcation method for urban air traffic scenarios. Background Art

[0002] With the rapid development of modern urban air traffic, the management of low-altitude airspace and the demarcation of air routes have become a key technical field to ensure the efficient and safe operation of aircraft. The current airspace division is mainly based on fixed air routes and regional management strategies, providing flight routes for aircraft in a static planning manner. With the diversification of application scenarios of drones and low-altitude aircraft, the airspace traffic density has shown a significant increase, and in order to cope with the possible more diverse air traffic of aircraft equipment, more sophisticated regulation is needed; Traditional methods often use fixed areas in two-dimensional or three-dimensional space to mark paths. They lack real-time analysis of multi-dimensional features such as aircraft density and flight patterns, making it difficult to respond flexibly to complex and changing flight environments. They are also not convenient for deep correlation of historical and current flight trajectory data, and are unable to form an effective trajectory mapping model. Path optimization and capacity prediction are often not carried out simultaneously, making it easy for the airspace to exceed the maximum load during busy hours. Paths cannot be adjusted in real time to adapt to the changing needs of aircraft density, affecting airspace traffic efficiency. Summary of the invention

[0003] In view of the problems existing in the above-mentioned existing low-altitude airspace route demarcation method for urban air traffic scenarios, the present invention is proposed.

[0004] Therefore, the problem to be solved by the present invention is that traditional methods often use fixed areas in two-dimensional or three-dimensional space to mark paths, lack real-time analysis methods for multi-dimensional features such as aircraft density and flight modes, and find it difficult to respond flexibly to complex and changeable flight environments. It is also inconvenient to deeply associate historical and current flight trajectory data, and it is impossible to form an effective trajectory mapping model, which affects the efficiency of airspace traffic.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a low-altitude airspace route design method for urban air traffic scenarios, which includes: Collect aircraft operation data in real time, subdivide aircraft flight altitude layers, and make preliminary urban airspace divisions; Based on the trajectory data, the aircraft trajectory sequence data is composed, and the trajectory sequence point mapping is performed. The DTW algorithm is used to calculate the similarity of the aircraft trajectory. The density peak clustering DPC algorithm is used to perform cluster analysis on the DTW similarity data to optimize the urban airspace division. The DDPG algorithm is used to build a deep learning model to predict the maximum number of aircraft and optimize the real-time path of aircraft; Sub-channel areas are delineated based on flight density, and sub-channel flight density is detected in real time for early warning.

[0006] As a preferred solution of the low-altitude airspace route demarcation method for urban air traffic scenarios described in the present invention, wherein: the real-time collection of aircraft operation data includes: Collect aircraft operation data in real time through CNS / ATM system, including position coordinates, speed, range, and flight time; Based on the operation data of different aircraft, the average flight speed, average flight altitude, flight frequency and flight density are calculated as flight characteristics.

[0007] As a preferred solution of the low-altitude airspace route demarcation method for urban air traffic scenarios described in the present invention, wherein: the subdividing the aircraft flight altitude layer and preliminarily dividing the urban airspace includes: Use pre-trained support vector machines (SVMs) to classify flight scenarios based on the flight characteristics of different aircraft; Based on each flight scenario, combined with the safety standards of the aircraft in the scenario, including the minimum flight altitude, the subdivision height of the altitude layer to which the aircraft belongs is determined, expressed as: ;in Indicates the recommended flight altitude for this type of aircraft. Indicates the lowest flight altitude starting from the ground, U indicates the airspace altitude layer to which the aircraft belongs according to safety standards, represents the average altitude of this type of aircraft, Indicates the minimum safe distance between aircraft; Determine the city center coordinates based on the position coordinate information and the take-off and landing coordinate information of the aircraft; The urban airspace is gridded through the Voronoi diagram, and the Voronoi diagram partitioning algorithm is used to perform preliminary regional division.

[0008] As a preferred solution of the low-altitude airspace route demarcation method for urban air traffic scenarios described in the present invention, wherein: based on the trajectory data, the aircraft trajectory sequence data is composed, the trajectory sequence point mapping is performed, the DTW algorithm is used to calculate the similarity of the aircraft trajectory, and the density peak clustering DPC algorithm is used to perform cluster analysis on the DTW similarity data, including: Collect the trajectory data of the aircraft and perform serial processing according to time, and form trajectory sequence data according to the positions of different aircraft at different times; According to the trajectory sequence data of different aircraft, the optimal path mapping table is constructed by dynamic programming DP method; The Euclidean distance between each point in the two trajectory sequences to be mapped is calculated to generate a distance matrix D; Construct the cumulative distance matrix C and use the dynamic programming algorithm to calculate the trajectory points arrive and arrive The shortest cumulative distance is expressed as: ; ; where n and m represent the number of trajectory points of trajectory a and trajectory b respectively, represents the shortest cumulative distance from the upper path, represents the shortest cumulative distance from the left path, represents the shortest cumulative distance from the diagonal path; From the end point of the cumulative distance matrix Start by backtracking along the minimum cumulative distance to find the optimal path mapping corresponding to the i-th node in the trajectory sequence ; Using the DTW algorithm, the similarity of two trajectories is calculated along the optimal path mapping, and the minimum distance value in the cumulative distance matrix is ​​used as the similarity, which is expressed as: ;in represents the similarity between two trajectory sequences a and b, n represents the total number of trajectory points in the trajectory sequence, and Represent the i-th trajectory point of trajectory sequences a and b respectively; The calculated DTW similarity results between the aircraft trajectories are organized into a similarity matrix, where N represents the number of aircraft, and cluster analysis is performed; The density peak clustering DPC algorithm is used to identify the trajectories with the highest density by performing density analysis on the DTW similarity matrix, and these density peak points are preferentially selected as the initial clustering centers, which can be expressed as: ;in represents the local density of the trajectory sequence, N represents the total number of all trajectories, Represents the mean parameter of the DTW distance based on all trajectory pairs; For each trajectory, count the trajectories with higher density and calculate the minimum distance between each trajectory and the trajectory with higher density, and identify the local maximum in the density peak as the initial cluster center, which is expressed as: ;in represents the minimum distance from each trajectory to the trajectory with higher density than it, and denote the local density of trajectories a and b respectively; Based on the initial cluster centers, each trajectory is assigned to the cluster center with the smallest DTW similarity according to the DTW similarity matrix; For each cluster, the trajectory sequence with the smallest average DTW distance from the intra-cluster trajectory to the new cluster center is recalculated as the new center. In each iteration, the cluster center is updated based on the similarity until the change in similarity is no longer obvious, then the iteration is stopped; Based on cluster analysis, the flight modes of different aircraft are identified and distinguished by iteratively selecting the trajectory with the smallest average DTW distance within the class as the new cluster center.

[0009] As a preferred solution of the low-altitude airspace route demarcation method for urban air traffic scenarios described in the present invention, wherein: the optimization of urban airspace division includes: Based on the grid division of urban airspace through Voronoi diagram, cluster centers are identified, and the main channel area is divided according to the trajectory sequence data of cluster centers. Based on the channel demarcation standard, if the distance between two cluster centers is less than the set minimum airspace interval, their corresponding Voronoi areas are merged to form a continuous airspace corridor.

[0010] As a preferred solution of the low-altitude airspace route demarcation method for urban air traffic scenarios described in the present invention, wherein: the DDPG algorithm is used to construct a deep learning model to predict the maximum number of aircraft and optimize the real-time path of the aircraft, including: The position coordinate data and velocity vector data of the aircraft and the relative distance from other aircraft are used as the state space data of the aircraft flight; The speed adjustment direction and size of the aircraft are used as action space data; Set a reward function, where the reward function considers the safe distance between aircraft, airspace congestion, and the smoothness of the flight path; The sum of the safety distance reward, airspace capacity reward and path smoothness reward is used as the total reward function; The DDPG algorithm is used to build a deep learning model, including the policy network Actor and the value network Critic for deep reinforcement learning training. The training set data is used to simulate a multi-aircraft environment in the airspace. The new position of the aircraft is determined according to the action space data and the next state space data is determined. The final reward is calculated based on the new state space data and the set reward function. Use the Bellman equation to calculate the target Q value, and use the target Q value calculated in the previous step to update it, and calculate the loss through the mean square error loss function; Iterative optimization is performed through the gradient descent method. When the calculated loss no longer changes significantly during the continuous iteration process, the iteration is stopped and the model parameters are output to update the model. Based on the clustering results, the real-time state space data and action space data of the aircraft are input to predict the maximum number of aircraft that can be accommodated in the current airspace, and real-time path planning is performed for each aircraft in the airspace.

[0011] As a preferred solution of the low-altitude airspace route demarcation method for urban air traffic scenarios described in the present invention, wherein: the sub-channel area is demarcated based on flight density, including: Based on the results of cluster analysis, the trajectory sequence along the high-density cluster center is used as the main path direction to calculate the flight density of the aircraft in the main path direction. The regional density threshold is set based on historical data. If the flight density exceeds the regional density threshold, sub-channel areas are designated on both sides of the main path direction.

[0012] As a preferred solution of the low-altitude airspace route demarcation method for urban air traffic scenarios described in the present invention, wherein: the real-time detection of sub-channel flight density for early warning includes: If a new aircraft enters the airspace, the flight density of the sub-channels on both sides is calculated in real time. With the goal of minimizing the difference in flight density between the sub-channels on both sides, the aircraft is dynamically assigned to different sub-channels. At the same time, through cluster analysis and DDPG algorithm deep learning model optimization, based on the sub-channel flight density and referring to the main path area density threshold, an alarm is issued if the sub-channel flight density exceeds the area density threshold.

[0013] A computer device comprises: a memory and a processor; the memory stores a computer program, and the processor implements the steps of the above-mentioned method for low-altitude airspace channel demarcation for urban air traffic scenarios when executing the computer program.

[0014] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for low-altitude airspace channel demarcation for urban air traffic scenarios.

[0015] The beneficial effects of the present invention are as follows: by performing density analysis on the DTW similarity matrix through DPC clustering, the density peak in the trajectory is found as the clustering center, forming a highly flexible clustering mechanism, and through further airspace division, efficient airspace division is achieved. Based on the clustering analysis results and the input state space data, the DDPG model can dynamically predict the maximum number of aircraft that can be accommodated in the current airspace, and optimize the real-time path planning of the aircraft, so that managers can allocate aircraft paths based on real-time capacity prediction data to prevent overcapacity risks. The optimized path planning reduces the waiting time of the aircraft and improves the traffic efficiency of the airspace. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0017] Figure 1 This is a flowchart of the low-altitude airspace route demarcation method for urban air traffic scenarios.

[0018] Figure 2 Schematic diagram of the optimized urban airspace division process for the low-altitude airspace route designation method for urban air traffic scenarios. DETAILED DESCRIPTION

[0019] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0020] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0021] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0022] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and which provides a low-altitude airspace channel demarcation method for urban air traffic scenarios. The low-altitude airspace channel demarcation method for urban air traffic scenarios includes: S1, collects aircraft operation data in real time, subdivides aircraft flight altitude layers, and preliminarily divides urban airspace; Preferably, real-time collection of aircraft operation data includes: Collect aircraft operation data in real time through CNS / ATM systems (communication, navigation, surveillance / air traffic management), including position coordinates, speed, range, and flight time; Based on the operation data of different aircraft, the average flight speed, average flight altitude, flight frequency and flight density are calculated as flight characteristics.

[0023] The CNS / ATM system can obtain the precise position, speed, range and flight time of an aircraft in real time, making it possible to accurately track the status of aircraft in the airspace. Real-time monitoring can not only promptly identify potential safety hazards, but also provide early warnings when the aircraft deviates from the planned path or enters an abnormal state. By calculating characteristic data such as the average speed, average altitude, flight frequency and flight density of the aircraft, we can accurately understand the operating rules of different flight modes and aircraft.

[0024] Furthermore, the flight altitude layers of aircraft are subdivided, and the urban airspace is initially divided, including: Use pre-trained support vector machines (SVMs) to classify flight scenarios based on the flight characteristics of different aircraft; Based on each flight scenario, combined with the safety standards of the aircraft in the scenario, including the minimum flight altitude, the subdivision height of the altitude layer to which the aircraft belongs is determined, expressed as: ;in Indicates the recommended flight altitude for this type of aircraft. Indicates the lowest flight altitude starting from the ground, U indicates the airspace altitude layer to which the aircraft belongs according to safety standards, represents the average altitude of this type of aircraft, Indicates the minimum safe distance between aircraft; Determine the city center coordinates based on the position coordinate information and the take-off and landing coordinate information of the aircraft; The urban airspace is gridded by the Voronoi diagram, and the Voronoi diagram partitioning algorithm is used to preliminarily divide the area, which is expressed as: ;in represents the i-th Voronoi region, Represents the two-dimensional plane space of the city, Represents point x to the i-th center point The Euclidean distance of Indicates that for all other center points except i Established.

[0025] By classifying the flight characteristics of different aircraft based on support vector machines (SVM), aircraft can be divided into different flight scenarios (such as logistics flights, emergency flights, etc.). Each type of flight scenario has specific flight behavior characteristics and requirements, which meet the needs of different flight modes and improve airspace utilization and safety. By dividing the altitude levels of aircraft based on the safety standards of each flight scenario (such as the minimum flight altitude, recommended flight altitude layer, etc.), spatial isolation between different aircraft is achieved, altitude layer conflicts are prevented, the complexity of airspace management is reduced, and flight safety is guaranteed. The coordinates of the city center are calculated based on the take-off and landing point data of the aircraft, so that the center of the airspace division conforms to the actual distribution of aircraft, which is conducive to channel design and airspace division. The city center is used as the division basis of the Voronoi diagram to ensure that the airspace division can cover the main flight activity area, improve the scientific nature of the airspace division, make airspace management and channel design meet actual needs, and ensure that the airspace division is reasonable and the coverage is effective. The division of Voronoi regions combined with the SVM classification results can realize the allocation and dynamic management of aircraft in each region. Combined with the DDPG algorithm for path planning and dynamic adjustment, the aircraft flow in each Voronoi region can be controlled within a reasonable range, and reasonable allocation of airspace resources can be achieved in high-density flight areas to avoid congestion, achieve safe and efficient management of airspace, and improve the overall utilization of airspace.

[0026] S2, based on the trajectory data, composes the aircraft trajectory sequence data, performs trajectory sequence point mapping, uses the DTW algorithm to calculate the similarity of the aircraft trajectory, adopts the density peak clustering DPC algorithm, performs cluster analysis on the DTW similarity data, and optimizes the urban airspace division; Preferably, based on the trajectory data, aircraft trajectory sequence data is formed, trajectory sequence point mapping is performed, the similarity of the aircraft trajectory is calculated using the DTW algorithm, and the density peak clustering DPC algorithm is used to perform cluster analysis on the DTW similarity data, including: Collect the trajectory data of the aircraft and perform serial processing according to time, and form trajectory sequence data according to the positions of different aircraft at different times; According to the trajectory sequence data of different aircraft, the optimal path mapping table is constructed by dynamic programming DP method; The Euclidean distance between each point in the two trajectory sequences to be mapped is calculated to generate a distance matrix D; Construct the cumulative distance matrix C and use the dynamic programming algorithm to calculate the trajectory points arrive and arrive The shortest cumulative distance is expressed as: ; ; where n and m represent the number of trajectory points of trajectory a and trajectory b respectively, represents the shortest cumulative distance from the upper path, represents the shortest cumulative distance from the left path, represents the shortest cumulative distance from the diagonal path; From the end point of the cumulative distance matrix Start by backtracking along the minimum cumulative distance to find the optimal path mapping corresponding to the i-th node in the trajectory sequence , expressed as: ; Using the DTW algorithm, the similarity of two trajectories is calculated along the optimal path mapping, and the minimum distance value in the cumulative distance matrix is ​​used as the similarity, which is expressed as: ;in represents the similarity between two trajectory sequences a and b, which is used to measure the dynamic time warping distance between two trajectory sequences. n represents the total number of trajectory points in the trajectory sequence. and Represent the i-th trajectory point of trajectory sequences a and b respectively; The calculated DTW similarity results between the aircraft trajectories are organized into a similarity matrix, where N represents the number of aircraft, and cluster analysis is performed; The density peak clustering DPC algorithm is used to identify the trajectories with the highest density by performing density analysis on the DTW similarity matrix, and these density peak points are preferentially selected as the initial clustering centers, which can be expressed as: ;in represents the local density of the trajectory sequence, N represents the total number of all trajectories, Represents the mean parameter of the DTW distance based on all trajectory pairs; For each trajectory, count the trajectories with higher density and calculate the minimum distance between each trajectory and the trajectory with higher density, and identify the local maximum in the density peak as the initial cluster center, which is expressed as: ;in represents the minimum distance from each trajectory to the trajectory with higher density than it, and denote the local density of trajectories a and b respectively; Based on the initial cluster center, each trajectory is assigned to the cluster center with the smallest DTW similarity according to the DTW similarity matrix, which is expressed as: ;in represents the clustering category described by trajectory sequence a, represents the jth cluster center; For each cluster, the trajectory sequence with the smallest average DTW distance from the intra-cluster trajectory to the new cluster center is recalculated as the new center. In each iteration, the cluster center is updated based on the similarity until the change in similarity is no longer obvious, then the iteration is stopped; Based on cluster analysis, the flight modes of different aircraft are identified and distinguished by iteratively selecting the trajectory with the smallest average DTW distance within the class as the new cluster center.

[0027] By collecting and serializing the trajectory data of the aircraft, the motion trajectory data of the aircraft is processed in time series to construct a detailed trajectory sequence, laying the foundation for subsequent analysis and classification. The dynamic programming algorithm is used to calculate the shortest cumulative distance from point to point in the trajectory sequence and construct an optimal path mapping table. This effectively solves the problem of time alignment in the trajectory data. Even if the trajectory durations of two aircraft are different, the corresponding trajectory points can be obtained through the optimal path mapping. By calculating the DTW similarity of two trajectory sequences, the similarity of aircraft trajectories can be evaluated on the basis of dynamic time warping to ensure that the time misalignment of the trajectories will not affect the matching quality. The DPC algorithm is used to perform density analysis on the DTW similarity matrix and identify the trajectory points with the highest density as the initial clustering center. Cluster analysis can be used to classify aircraft trajectories with high similarity to achieve aircraft pattern recognition. Different aircraft groups and flight modes can be quickly and accurately identified, and aircraft with similar characteristics can be clustered in the same category, which is helpful for airspace management and scheduling optimization. In the clustering process, the center point of each cluster is iteratively updated to ensure that the center point of each class always represents the average characteristics within the class. The trajectory sequence with the smallest average DTW distance within the class is used as the new cluster center to gradually optimize the clustering results, effectively improving the precision and accuracy of the clustering results, ensuring that the clustering results can accurately reflect the different flight modes of the aircraft, and achieving more accurate airspace monitoring and behavior analysis. Traditional similarity calculation methods are difficult to deal with irregular trajectories and the time-space misalignment between different aircraft. However, by optimizing trajectory mapping through dynamic programming and then using DTW to calculate similarity, it is not only possible to effectively align trajectories of different lengths, but also to accurately calculate similarity based on the flight characteristics of different aircraft. This method has unique applicability and advantages in processing asymmetric and time-misaligned trajectory data. After completing the calculation of trajectory similarity, DPC clustering performs density analysis on the DTW similarity matrix and finds the density peak in the trajectory as the clustering center. Traditional clustering methods (such as K-means) require a preset number of classes, while DPC clustering adaptively discovers the flight mode centers of different classes through density. It is particularly suitable for airspace environments with sparse and dynamically changing aircraft distribution. Combining the density peak with the dynamic time warping distance not only significantly improves the clustering accuracy, but also can adapt to the characteristics of different flight modes to form a highly flexible clustering mechanism.

[0028] Further, optimize the urban airspace division, including: Based on the grid division of urban airspace through Voronoi diagram, cluster centers are identified, and the main channel area is divided according to the trajectory sequence data of cluster centers. Based on the channel demarcation standard, if the distance between two cluster centers is less than the set minimum airspace interval, their corresponding Voronoi areas are merged to form a continuous airspace corridor.

[0029] Through further airspace division, efficient airspace division is achieved, ensuring the efficiency of each channel, reasonably allocating airspace resources, making channel demarcation meet actual flight needs, reducing overlap between channels, and improving overall airspace management efficiency. By reasonably merging channels in dense areas, the main channel structure is flexible, avoiding crowded airspace environments, and achieving effective management of high-density areas. By merging adjacent Voronoi areas, the system can automatically demarcate coherent channels in high-density areas, reducing the complexity of managing multiple small areas. At the same time, the positioning of cluster centers helps to automatically generate channel layouts, making airspace demarcation more intelligent.

[0030] S3, uses the DDPG algorithm to build a deep learning model to predict the maximum number of aircraft and optimize the real-time path of aircraft; Preferably, a DDPG algorithm is used to construct a deep learning model to predict the maximum number of aircraft and optimize the real-time path of the aircraft, including: The position coordinate data and velocity vector data of the aircraft and the relative distance from other aircraft are used as the state space data of the aircraft flight; The speed adjustment direction and size of the aircraft are used as action space data; Set the reward function, where the reward function considers the safe distance between aircraft, airspace congestion, and the stability of the flight path, expressed as: ; in represents the Euclidean distance between aircraft i and j, M represents the total number of aircraft, represents the distance penalty coefficient, Indicates safe distance reward; ; in represents the airspace capacity bonus, Indicates the number of aircraft in the current airspace. Represents the area or volume of airspace. represents the airspace capacity target value, represents the capacity optimization factor; ;in represents the path smoothness reward, represents the total time step, represents the speed adjustment of the aircraft at time step t, represents the stability optimization coefficient; Based on historical flight data, it is screened according to safe flight standards to optimize data, including , and Determined based on optimization data respectively; The sum of the safety distance reward, airspace capacity reward and path smoothness reward is used as the total reward function; The DDPG algorithm is used to build a deep learning model, including the policy network Actor and the value network Critic for deep reinforcement learning training. The training set data is used to simulate a multi-aircraft environment in the airspace. The new position of the aircraft is determined according to the action space data and the next state space data is determined. The final reward is calculated based on the new state space data and the set reward function. The target Q value is calculated using the Bellman equation and updated using the target Q value calculated in the previous step. The loss is calculated using the mean square error loss function, expressed as: ;in Represents the computational loss between the current Q value estimate and the target Q value, Represents the model parameters, including the weights and biases in the network, represents the i-th state space data, Indicated in The execution action data under, B represents the total number of sample data, represents the target Q value; Iterative optimization is performed through the gradient descent method. When the calculated loss no longer changes significantly during the continuous iteration process, the iteration is stopped and the model parameters are output to update the model. Based on the clustering results, the real-time state space data and action space data of the aircraft are input to predict the maximum number of aircraft that can be accommodated in the current airspace, and real-time path planning is performed for each aircraft in the airspace.

[0031] When executing real-time path planning, the DDPG model first receives the classification information of aircraft behavior patterns from cluster analysis. Based on the current aircraft status information (position, speed, etc.) and cluster categories, the DDPG model generates predicted optimal path planning actions (adjusting speed and direction) for the aircraft through the policy network (Actor) and value network (Critic) contained in the model based on the policy network and cluster analysis information. As the aircraft status continues to change, the DDPG model will adjust the aircraft's path in real time to ensure a safe distance while improving flight efficiency. During the flight, DDPG will continuously optimize the path planning results according to the reward mechanism of different cluster categories, so that the optimization of the aircraft path can be further affected by optimizing and adjusting the reward function. At the same time, the value of the current state and action is evaluated based on the value network, and the path planning and density distribution are jointly evaluated in combination with the reward function to determine the maximum number of aircraft that can be accommodated. By continuously adjusting the flight path and speed of the aircraft, optimizing the flight path, and improving the utilization rate of airspace, dynamic path optimization enables aircraft to automatically avoid crowded areas and potential risk areas, and realize efficient use of resources in the airspace. By setting rewards, the safety distance reward improves the safety between aircraft and reduces the potential risk of collision. The airspace capacity reward optimizes the distribution of aircraft in the airspace, prevents the aircraft from being too dense in a specific area, improves the management efficiency of flight traffic, realizes balanced airspace resource allocation, reduces the acceleration and turning frequency of aircraft, and improves the flight stability and energy efficiency of aircraft. The target Q value and mean square error loss are calculated through the Bellman equation, and the DDPG model is trained to find the optimal strategy between states and actions, thereby maximizing the path stability and airspace management efficiency of the aircraft, achieving effective path planning and capacity prediction for complex airspace, improving training efficiency, and ensuring that the model can respond quickly in practical applications; Based on the cluster analysis results and the input state space data, the DDPG model can dynamically predict the maximum number of aircraft that can be accommodated in the current airspace and optimize the real-time path planning of aircraft, enabling managers to allocate aircraft paths based on real-time capacity prediction data to prevent overcapacity risks. Optimized path planning reduces aircraft waiting time and improves airspace traffic efficiency.

[0032] S4, demarcate sub-channel areas based on flight density, detect sub-channel flight density in real time and issue early warnings; Preferably, sub-channel areas are defined based on flight density, including: Based on the results of cluster analysis, the trajectory sequence along the high-density cluster center is used as the main path direction to calculate the flight density of the aircraft in the main path direction. The regional density threshold is set based on historical data. If the flight density exceeds the regional density threshold, sub-channel areas are designated on both sides of the main path direction.

[0033] Sub-channels can effectively alleviate congestion in main channels, reduce delays caused by high density of aircraft, ensure smooth passage of aircraft in the airspace, improve the utilization efficiency of airspace resources, improve the management efficiency of airspace, and avoid channel interference caused by speed or mode differences of aircraft.

[0034] Furthermore, the sub-channel flight density is detected in real time for early warning, including: If a new aircraft enters the airspace, the flight density of the sub-channels on both sides is calculated in real time. With the goal of minimizing the difference in flight density between the sub-channels on both sides, the aircraft is dynamically assigned to different sub-channels. At the same time, through cluster analysis and DDPG algorithm deep learning model optimization, based on the sub-channel flight density and referring to the main path area density threshold, an alarm is issued if the sub-channel flight density exceeds the area density threshold.

[0035] By calculating the flight density difference of the sub-channels on both sides, new aircraft are assigned to the channel with lower density, with the goal of minimizing the density difference, ensuring balanced traffic distribution when new aircraft enter, and preventing congestion in a single channel. When the flight density of the sub-channel exceeds the set main path area density threshold, an alarm is automatically issued to remind airspace management personnel to regulate traffic or guide some aircraft to adjust to other channels, effectively avoiding the safety hazards caused by excessively dense flight, ensuring safe intervals between aircraft, and reducing the risk of collision between channels. Example

[0036] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0037] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0038] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0039] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit with a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit with a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0040] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A low-altitude airspace route designation method for urban air traffic scenarios, characterized by: include, Collect aircraft operation data in real time, subdivide aircraft flight altitude layers, and make preliminary urban airspace divisions; Based on the trajectory data, the aircraft trajectory sequence data is composed, and the trajectory sequence point mapping is performed. The DTW algorithm is used to calculate the similarity of the aircraft trajectory. The density peak clustering DPC algorithm is used to perform cluster analysis on the DTW similarity data to optimize the urban airspace division. The DDPG algorithm is used to build a deep learning model to predict the maximum number of aircraft and optimize the real-time path of aircraft; Sub-channel areas are delineated based on flight density, and sub-channel flight density is detected in real time for early warning.

2. The low-altitude airspace route designation method for urban air traffic scenarios as claimed in claim 1 is characterized by: The real-time collection of aircraft operation data includes: Collect aircraft operation data in real time through CNS / ATM system, including position coordinates, speed, range, and flight time; Based on the operation data of different aircraft, the average flight speed, average flight altitude, flight frequency and flight density are calculated as flight characteristics.

3. The low-altitude airspace route designation method for urban air traffic scenarios as claimed in claim 2 is characterized by: The subdivision of aircraft flight altitude layers and the preliminary division of urban airspace include: Use pre-trained support vector machines (SVMs) to classify flight scenarios based on the flight characteristics of different aircraft; Based on each flight scenario, combined with the safety standards of the aircraft in the scenario, including the minimum flight altitude, the subdivision height of the altitude layer to which the aircraft belongs is determined, expressed as: ;in Indicates the recommended flight altitude for this type of aircraft. Indicates the lowest flight altitude starting from the ground, U indicates the airspace altitude layer to which the aircraft belongs according to safety standards, represents the average altitude of this type of aircraft, Indicates the minimum safe distance between aircraft; Determine the city center coordinates based on the position coordinate information and the take-off and landing coordinate information of the aircraft; The urban airspace is gridded through the Voronoi diagram, and the Voronoi diagram partitioning algorithm is used to perform preliminary regional division.

4. The low-altitude airspace route designation method for urban air traffic scenarios as claimed in claim 3 is characterized by: Based on the trajectory data, the aircraft trajectory sequence data is composed, the trajectory sequence point mapping is performed, the DTW algorithm is used to calculate the similarity of the aircraft trajectory, and the density peak clustering DPC algorithm is used to perform cluster analysis on the DTW similarity data, including: Collect the trajectory data of the aircraft and perform serial processing according to time, and form trajectory sequence data according to the positions of different aircraft at different times; According to the trajectory sequence data of different aircraft, the optimal path mapping table is constructed by dynamic programming DP method; The Euclidean distance between each point in the two trajectory sequences to be mapped is calculated to generate a distance matrix D; Construct the cumulative distance matrix C and use the dynamic programming algorithm to calculate the trajectory points arrive and arrive The shortest cumulative distance is expressed as: ; ; where n and m represent the number of trajectory points of trajectory a and trajectory b respectively, represents the shortest cumulative distance from the upper path, represents the shortest cumulative distance from the left path, represents the shortest cumulative distance from the diagonal path; From the end point of the cumulative distance matrix Start by backtracking along the minimum cumulative distance to find the optimal path mapping corresponding to the i-th node in the trajectory sequence ; Using the DTW algorithm, the similarity of two trajectories is calculated along the optimal path mapping, and the minimum distance value in the cumulative distance matrix is ​​used as the similarity, which is expressed as: ;in represents the similarity between two trajectory sequences a and b, n represents the total number of trajectory points in the trajectory sequence, and Represent the i-th trajectory point of trajectory sequences a and b respectively; The calculated DTW similarity results between the aircraft trajectories are organized into a similarity matrix, where N represents the number of aircraft, and cluster analysis is performed; The density peak clustering DPC algorithm is used to identify the trajectories with the highest density by performing density analysis on the DTW similarity matrix, and these density peak points are preferentially selected as the initial clustering centers, which can be expressed as: ;in represents the local density of the trajectory sequence, N represents the total number of all trajectories, Represents the mean parameter of the DTW distance based on all trajectory pairs; For each trajectory, count the trajectories with higher density and calculate the minimum distance between each trajectory and the trajectory with higher density, and identify the local maximum in the density peak as the initial cluster center, which is expressed as: ;in represents the minimum distance from each trajectory to the trajectory with higher density than it, and denote the local density of trajectories a and b respectively; Based on the initial cluster centers, each trajectory is assigned to the cluster center with the smallest DTW similarity according to the DTW similarity matrix; For each cluster, the trajectory sequence with the smallest average DTW distance from the intra-cluster trajectory to the new cluster center is recalculated as the new center. In each iteration, the cluster center is updated based on the similarity until the change in similarity is no longer obvious, then the iteration is stopped; Based on cluster analysis, the flight modes of different aircraft are identified and distinguished by iteratively selecting the trajectory with the smallest average DTW distance within the class as the new cluster center.

5. The low-altitude airspace route designation method for urban air traffic scenarios as claimed in claim 4 is characterized by: The optimization of urban airspace division includes: Based on the grid division of urban airspace through Voronoi diagram, cluster centers are identified, and the main channel area is divided according to the trajectory sequence data of cluster centers. Based on the channel demarcation standard, if the distance between two cluster centers is less than the set minimum airspace interval, their corresponding Voronoi areas are merged to form a continuous airspace corridor.

6. The low-altitude airspace route designation method for urban air traffic scenarios as claimed in claim 5 is characterized by: The DDPG algorithm is used to construct a deep learning model to predict the maximum number of aircraft and optimize the real-time path of the aircraft, including: The position coordinate data and velocity vector data of the aircraft and the relative distance from other aircraft are used as the state space data of the aircraft flight; The speed adjustment direction and size of the aircraft are used as action space data; Set a reward function, where the reward function considers the safe distance between aircraft, airspace congestion, and the smoothness of the flight path; The sum of the safety distance reward, airspace capacity reward and path smoothness reward is used as the total reward function; The DDPG algorithm is used to build a deep learning model, including the policy network Actor and the value network Critic for deep reinforcement learning training. The training set data is used to simulate a multi-aircraft environment in the airspace. The new position of the aircraft is determined according to the action space data and the next state space data is determined. The final reward is calculated based on the new state space data and the set reward function. Use the Bellman equation to calculate the target Q value, and use the target Q value calculated in the previous step to update it, and calculate the loss through the mean square error loss function; Iterative optimization is performed through the gradient descent method. When the calculated loss no longer changes significantly during the continuous iteration process, the iteration is stopped and the model parameters are output to update the model. Based on the clustering results, the real-time state space data and action space data of the aircraft are input to predict the maximum number of aircraft that can be accommodated in the current airspace, and real-time path planning is performed for each aircraft in the airspace.

7. The low-altitude airspace route designation method for urban air traffic scenarios as claimed in claim 6 is characterized by: The sub-channel area is defined based on flight density, including: Based on the results of cluster analysis, the trajectory sequence along the high-density cluster center is used as the main path direction to calculate the flight density of the aircraft in the main path direction. The regional density threshold is set based on historical data. If the flight density exceeds the regional density threshold, sub-channel areas are designated on both sides of the main path direction.

8. The low-altitude airspace route designation method for urban air traffic scenarios as claimed in claim 7 is characterized by: The real-time detection of sub-channel flight density is used for early warning. include, If a new aircraft enters the airspace, the flight density of the sub-channels on both sides is calculated in real time. With the goal of minimizing the difference in flight density between the sub-channels on both sides, the aircraft is dynamically assigned to different sub-channels. At the same time, through cluster analysis and DDPG algorithm deep learning model optimization, based on the sub-channel flight density and referring to the main path area density threshold, an alarm is issued if the sub-channel flight density exceeds the area density threshold.

9. A computer device comprising: Memory and processor; The memory stores a computer program, which is characterized in that when the processor executes the computer program, the steps of the low-altitude airspace channel demarcation method for urban air traffic scenarios described in any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the low-altitude airspace channel demarcation method for urban air traffic scenarios described in any one of claims 1 to 8 are implemented.

Images