Airport digital taxiway map modeling method and system based on flight trajectory data

Abstract

The present application relates to the technical field of traffic information mining, in particular to an airport digital taxiway map modeling method and system based on flight trajectory data, comprising data preprocessing of flight trajectory, establishment of flight apron time sequence trajectory subsequence dataset; vertical projection distance and angle projection determination are performed on the flight apron time sequence subsequence dataset, and potential straight section points and curve section points are identified; the subsequence set division time window is adjusted, double determination is performed on the new subsequence dataset, shape boundary processing is performed on the detected conflict section, and the determined straight line section and curve section are divided; the straight line section set and the curve section set are incrementally clustered, and the path cluster credible endpoint set is identified; the path credible endpoint set is fused, and the path structure is output, solving the problem that it is difficult to automatically and accurately extract and construct an airport digital taxiway network from massive discrete actual flight taxiway trajectories depending on manual surveying and mapping or airport drawings.

Description

Technical Field

[0001] This invention relates to the field of traffic information mining technology, and in particular to a method and system for modeling airport digital taxiway maps based on flight trajectory data. Background Technology

[0002] As a critical node in the air transport system, the efficiency and safety of airport operations are paramount. Detailed, high-fidelity airport digital maps are a core foundation for enhancing airport operational control capabilities. Digital maps play an irreplaceable supporting role in advanced applications such as taxiing control, conflict warning, route planning, resource scheduling, and operational simulation. However, constructing a digital map that accurately reflects the actual taxiing rules and spatial structure of aircraft still faces significant challenges.

[0003] Existing methods for constructing digital models of airport taxiways, through algorithms such as trajectory clustering and centerline extraction, can reflect the actual operating patterns of aircraft. However, due to the characteristics of airport surface trajectory data, such as high noise, uneven sampling rate, a large number of waiting and stopping points, and mixed features of straight and turning segments, existing algorithms generally suffer from problems such as loss of details, inaccurate curve identification, and difficulty in extracting road network topology (such as intersections and endpoints). This results in broken, redundant, or geometrically deformed path networks in the generated paths. Therefore, manual methods cannot adapt to the rapidly changing operating environment and the needs of big data analysis. As for automated methods based on trajectory data, there is still a lack of effective solutions for how to robustly and accurately extract continuous, complete, and structurally clear standardized taxiway networks from discrete and non-uniform surface taxiways. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for modeling airport digital taxiway maps based on flight trajectory data, which can automatically and accurately reconstruct the airport taxiway network from discrete and discontinuous actual flight trajectory data through the provided system and method.

[0005] To achieve the above objectives, this invention provides a method for modeling airport digital taxiway maps based on flight trajectory data, comprising the following steps:

[0006] Flight trajectory data preprocessing is performed to establish a flight scene time-series trajectory subsequence dataset;

[0007] By performing dual determination of vertical projection distance and angular projection on the flight scene time sequence subsequence dataset, potential straight-line segments and curve segments are identified;

[0008] Adjust the time window for subsequence set division, perform double judgment again on the new subsequence dataset, and process the shape boundary of the detected conflict segments to divide the straight line segments and curved segments into defined segments.

[0009] Incremental clustering is performed on the straight line segment set and the curved line segment set respectively to identify the credible endpoint set of the path cluster;

[0010] Endpoint fusion is performed on the set of trusted endpoints of the path to output the airport taxiway network structure consisting of nodes and path edges.

[0011] Specifically, the process of preprocessing flight trajectories to establish a flight scene time-series trajectory subsequence dataset includes:

[0012] The original trajectory data is grouped and sorted based on the unique identifier of the aircraft to form a time-series trajectory set for each aircraft.

[0013] By combining the airport surface area range, spatial filtering of trajectory points is performed to extract surface taxiing trajectory sequences;

[0014] By setting the starting position and length of the subsequence, the scene gliding trajectory sequence is divided into equal-length segments to construct a time-sequence subsequence dataset.

[0015] Specifically, the dual determination of vertical projection distance and angular projection on the flight scene time-series sub-sequence dataset includes:

[0016] For each subsequence, construct the main direction vector based on its starting and ending points;

[0017] Calculate the vertical projection distance of each intermediate trajectory point within the subsequence relative to the main direction vector, and the angle between the direction vector of that point and the main direction vector;

[0018] Based on the set vertical projection distance threshold and angle threshold, the subsequence is determined to be a straight segment or a curved segment, and the internal trajectory point type is marked accordingly.

[0019] Specifically, the adjustment of the subsequence set to divide the time window and the shape boundary processing of the conflict segment include:

[0020] By changing the starting position of the subsequence, the new subsequence dataset is subjected to double determination again, resulting in two configuration determination results for each trajectory point;

[0021] The set of continuous trajectory points whose two judgment results are inconsistent is defined as the conflict segment;

[0022] The conflict segment is divided into two parts, front and back, based on the midpoint of the conflict segment, and then assigned to the straight segment set and the curve segment set respectively, thus completing the stable structured segmentation.

[0023] Specifically, the incremental clustering of the straight segment set and the curved segment set and the identification of the reliable endpoint set of the path cluster include:

[0024] Define geometric parameters for the set of straight segments and the set of curved segments, including the start point, end point, and direction or curvature characteristics;

[0025] Incremental clustering is performed on straight segments according to the trajectory sequence order, and straight segments that meet the requirements of directional consistency, spatial proximity and main direction continuity are merged into straight path clusters;

[0026] Extract the endpoints of all straight segments within the straight path cluster and perform density-based spatial clustering. Select the endpoint sub-clusters with the smallest and largest projections along the main direction as the reliable starting point and reliable ending point.

[0027] The same method is applied to the set of curve segments to obtain the curve path cluster and its reliable endpoints.

[0028] Specifically, the endpoint fusion of the trusted endpoint set of the path and the output path structure include:

[0029] By introducing a spatial location threshold, endpoints of straight path clusters and curved path clusters whose Euclidean distance is not greater than the threshold are merged into the same spatial node, forming a unified set of path boundary nodes.

[0030] Construct a straight sliding edge between the start and end nodes of the straight path cluster, and construct a turning sliding edge between the start and end nodes of the curved path cluster.

[0031] The airport taxiway network structure is composed of the set of nodes and the set of path edges.

[0032] The unique identifier of an aircraft includes the aircraft number or flight number;

[0033] The time-series trajectory set is obtained by sorting it in ascending order by timestamp;

[0034] The airport surface area is a polygonal structure region, and the spatial filtering of trajectory points is achieved through point-polygon inclusion detection.

[0035] The vertical projection distance is calculated by vector cross product and Euclidean norm.

[0036] The included angle is calculated using the vector dot product and the inverse cosine function;

[0037] If all intermediate trajectory points in a subsequence satisfy the condition that the vertical projection distance is not greater than the threshold and the included angle is not greater than the threshold, it is determined to be a straight segment; otherwise, it is determined to be a curved segment.

[0038] The geometric parameters of the straight segment include the start point, the end point, and the main direction angle;

[0039] The geometric parameters of the curved section include the starting point, the ending point, and the curvature characteristic parameters;

[0040] The density-based spatial clustering is implemented using the DBSCAN algorithm.

[0041] The airport digital taxiway map modeling system based on flight trajectory data is used to implement the airport digital taxiway map modeling method based on flight trajectory data. It is characterized by comprising a data preprocessing module, a trajectory segment identification module, a conflict segment processing module, a path clustering module, and a road network construction module. The trajectory segment identification module is connected to the data preprocessing module, the conflict segment processing module is connected to the trajectory segment identification module, the path clustering module is connected to the conflict segment processing module, and the road network construction module is connected to the path clustering module.

[0042] The data preprocessing module is used to preprocess flight trajectories and establish a flight scene time-series trajectory subsequence dataset.

[0043] The trajectory segmentation and recognition module is used to perform dual determination of vertical projection distance and angular projection on the flight scene time-series subsequence dataset to identify potential straight-line segments and curve segments.

[0044] The conflict segment processing module is used to adjust the time window for dividing the subsequence set, perform double judgment again on the new subsequence dataset, and process the shape boundary of the detected conflict segment to divide the line segment and curve segment into a defined line segment and curve segment.

[0045] The path clustering module is used to incrementally cluster the straight line segment set and the curved line segment set respectively, and identify the reliable endpoint set of the path cluster.

[0046] The road network construction module is used to perform endpoint fusion on the set of trusted endpoints of the path and output the airport taxiway network structure composed of nodes and path edges.

[0047] The present invention provides an airport digital taxiway map modeling method and system based on flight trajectory data. The method involves preprocessing flight trajectories to establish a time-series flight path sub-sequence dataset; determining the vertical projection distance and angular projection of the time-series flight path sub-sequence dataset to identify potential straight and curved segments; adjusting the time window for sub-sequence set division; performing dual determination on the new sub-sequence dataset; processing the shape boundaries of detected conflict segments to define definite straight and curved segments; incrementally clustering the straight and curved segments to identify a set of reliable endpoints for path clusters; and fusion of the reliable endpoints to output the path structure. This method aims to solve the problem in existing technologies that rely on manual surveying or airport maps, making it difficult to automatically and accurately extract and construct airport digital taxiway networks from massive amounts of discrete actual flight taxiways. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0049] Figure 1 This is a flowchart of the airport digital taxiway map modeling method based on flight trajectory data according to the present invention.

[0050] Figure 2 This is a schematic diagram of the vertical projection distance and angle projection determination of the present invention.

[0051] Figure 3 This is a schematic diagram of DBSCAN spatial reliable point clustering of the present invention.

[0052] Figure 4 This is a schematic diagram of the structure of the airport digital taxiway map modeling system based on flight trajectory data of the present invention.

[0053] In the diagram: 1-Data preprocessing module, 2-Trajectory segmentation and identification module, 3-Conflict segment processing module, 4-Path clustering module, 5-Road network construction module. Detailed Implementation

[0054] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0055] In the description of this invention, it should be understood that "a plurality of" means two or more, unless otherwise explicitly specified.

[0056] Please see Figures 1 to 3 This invention provides a method for modeling airport digital taxiway maps based on flight trajectory data, comprising the following steps:

[0057] S11: Perform data preprocessing on flight trajectories to establish a dataset of flight scene time-series trajectory subsequences;

[0058] S12: Perform dual judgment of vertical projection distance and angular projection on the flight scene time sequence subsequence dataset to identify potential straight-line segments and curve segments;

[0059] S13: Adjust the time window for subsequence set division, perform double judgment again on the new subsequence dataset, and process the shape boundary of the detected conflict segments to divide the straight line segments and curved segments.

[0060] S14: Perform incremental clustering on the straight line segment set and the curved line segment set respectively to identify the credible endpoint set of the path cluster;

[0061] Endpoint fusion is performed on the set of trusted endpoints of the path to output the airport taxiway network structure consisting of nodes and path edges.

[0062] Furthermore, the process of preprocessing flight trajectories to establish a flight scene time-series trajectory subsequence dataset specifically includes:

[0063] The original trajectory data is grouped and sorted based on the unique identifier of the aircraft to form a time-series trajectory set for each aircraft.

[0064] By combining the airport surface area range, spatial filtering of trajectory points is performed to extract surface taxiing trajectory sequences;

[0065] By setting the starting position and length of the subsequence, the scene gliding trajectory sequence is divided into equal-length segments to construct a time-sequence subsequence dataset.

[0066] Furthermore, the unique aircraft identifier includes the aircraft number or flight number;

[0067] The time-series trajectory set is obtained by sorting it in ascending order by timestamp;

[0068] The airport surface area is a polygonal structure region, and the spatial filtering of trajectory points is achieved through point-polygon inclusion detection.

[0069] In this embodiment, the acquired original airport flight location trajectory dataset is set during use. , No. A trajectory point is defined as:

[0070] ;

[0071] in, A unique identifier for an aircraft, such as the aircraft number or flight number; For the timestamp of the trajectory point, These are the trajectory position coordinates in the airport's plane coordinate system. Based on aircraft identification. For trajectory datasets The data is grouped to obtain the trajectory data group for each aircraft. ,Right now:

[0072] ;

[0073] in, Indicates the first aircraft, For the number of aircraft.

[0074] For each aircraft dataset According to timestamp Sort in ascending order to obtain the time series trajectory set. ,in This indicates the number of trajectory points for the aircraft.

[0075] The airport surface area is defined as a polygonal structure. ,right Perform point-polygon inclusion detection on each trajectory point:

[0076] ;

[0077] Retain satisfaction The trajectory points are used to filter out trajectory sequences that match the scene's gliding motion. :

[0078] ;

[0079] Furthermore, regarding the scene trajectory sequence Subsequence partitioning is performed to construct a flight scene time-series trajectory subsequence dataset. (Settings...) This is the starting position of the subsequence, and the length of each subsequence is... Then the aircraft The The temporal subsequence of the scene trajectory can be represented as:

[0080] .

[0081] Furthermore, the dual determination of vertical projection distance and angular projection on the flight scene time-series subsequence dataset specifically includes:

[0082] For each subsequence, construct the main direction vector based on its starting and ending points;

[0083] Calculate the vertical projection distance of each intermediate trajectory point within the subsequence relative to the main direction vector, and the angle between the direction vector of that point and the main direction vector;

[0084] Based on the set vertical projection distance threshold and angle threshold, the subsequence is determined to be a straight segment or a curved segment, and the internal trajectory point type is marked accordingly.

[0085] Furthermore, the vertical projection distance is calculated using vector cross product and Euclidean norm;

[0086] The included angle is calculated using the vector dot product and the inverse cosine function;

[0087] If all intermediate trajectory points in a subsequence satisfy the condition that the vertical projection distance is not greater than the threshold and the included angle is not greater than the threshold, it is determined to be a straight segment; otherwise, it is determined to be a curved segment.

[0088] In this embodiment, when used, for sub-sequences Set the starting position of the subsequence =0, defining the starting trajectory point as Define the termination trajectory point .

[0089] This constructs a subsequence Main direction vector ,in, This indicates the overall taxiing direction of the aircraft within this subsequence.

[0090] For subsequence any intermediate trajectory point ,in ,like Figure 2 Define the vector from the trajectory point to the principal direction. The vertical projection distance is:

[0091] ;

[0092] symbol Represents the cross product of vectors, symbol Describing the Euclidean norm, This represents the vertical projection distance of the trajectory point relative to the main direction of the subsequence.

[0093] Simultaneously, construct trajectory points Direction vector , trajectory point direction vector With the main direction vector of the subsequence The included angle between them is defined as:

[0094] ;

[0095] Among them, symbols This represents the vector dot product operation.

[0096] Set the vertical projection distance threshold to The angle threshold is Then the subsequence The determination rule is as follows:

[0097] When all intermediate trajectory points within a subsequence satisfy: Then the subsequence is determined to be a straight segment, and all trajectory points within it are straight segment points;

[0098] If any trajectory point does not meet the above conditions, then the subsequence is determined to be a curve segment, and all trajectory points within it are curve segment points.

[0099] Each subsequence trajectory points The possible judgment result is: , representing a straight segment point; , which represents a curve segment.

[0100] Furthermore, the adjustment of the subsequence set to divide the time window and the shape boundary processing of conflict segments specifically includes:

[0101] By changing the starting position of the subsequence, the new subsequence dataset is subjected to double determination again, resulting in two configuration determination results for each trajectory point;

[0102] The set of continuous trajectory points whose two judgment results are inconsistent is defined as the conflict segment;

[0103] The conflict segment is divided into two parts, front and back, based on the midpoint of the conflict segment, and then assigned to the straight segment set and the curve segment set respectively, thus completing the stable structured segmentation.

[0104] In this embodiment, considering the potential configuration misjudgment of aircraft timing subsequences at turning points, the time window for dividing the subsequence set is adjusted, and the starting position of the subsequence is set. = By performing dual determination on the new subsequence dataset, each trajectory point... Configuration determination results It can be represented as:

[0105] ;

[0106] When satisfied Define the trajectory point as a conflict segment, and define the set of trajectories composed of the consecutive conflicting trajectory points as the first... A conflict segment is represented as:

[0107] ;

[0108] in, and These represent the start and end indices of the conflict segment in the trajectory sequence, respectively.

[0109] Definition of the first The number of trajectory points contained in each conflict segment is: The split index position of the conflict segment is further defined as follows:

[0110] ;

[0111] Among them, symbols This indicates the floor function.

[0112] According to the trajectory points in the time series The order of the conflict segments Divided into two parts: front and back.

[0113] ;

[0114] ;

[0115] Considering the temporal continuity and spatial transition characteristics between straight and curved sections during aircraft taxiing, the first and second halves of the conflict segment are respectively assigned to the straight section set and the curved section set, as follows:

[0116] ;

[0117] in: Indicates aircraft The set of straight segments; Indicates aircraft A collection of curved sections.

[0118] After the conflict segment allocation process is completed, the aircraft Scene gliding trajectory sequence It is decomposed into a set of straight segments and a set of curved segments, satisfying:

[0119] ;

[0120] This enables stable, structured segmentation of aircraft surface taxiing trajectory sequences.

[0121] Furthermore, the incremental clustering of the straight segment set and the curved segment set and the identification of the reliable endpoint set of the path cluster specifically includes:

[0122] Define geometric parameters for the set of straight segments and the set of curved segments, including the start point, end point, and direction or curvature characteristics;

[0123] Incremental clustering is performed on straight segments according to the trajectory sequence order, and straight segments that meet the requirements of directional consistency, spatial proximity and main direction continuity are merged into straight path clusters;

[0124] Extract the endpoints of all straight segments within the straight path cluster and perform density-based spatial clustering. Select the endpoint sub-clusters with the smallest and largest projections along the main direction as the reliable starting point and reliable ending point.

[0125] The same method is applied to the set of curve segments to obtain the curve path cluster and its reliable endpoints.

[0126] Furthermore, the geometric parameters of the straight segment include the start point, the end point, and the main direction angle;

[0127] The geometric parameters of the curved section include the starting point, the ending point, and the curvature characteristic parameters;

[0128] The density-based spatial clustering is implemented using the DBSCAN algorithm.

[0129] In this embodiment, the surface trajectories of all aircraft are processed to obtain a set of straight-line segments. Combined with the curve section For the set of straight segments Each straight segment Its geometric parameters are defined as follows:

[0130] ;

[0131] in, These are the start and end points of the straight section, respectively. The main direction angle is determined by the start and end points.

[0132] Collection of curve segments Each curve section Its geometric parameters are defined as follows:

[0133] ;

[0134] in, These are the start and end points of the curve section, respectively. The curvature characteristic parameter is used to characterize the shape of a curve.

[0135] First, incremental clustering is performed on the straight segments sequentially according to their order in the trajectory sequence. For the current straight segment... When it is combined with an existing cluster of straight paths It meets the preset judgment conditions in terms of directional consistency, spatial proximity, and continuity along the main direction. If the condition is not met, the straight segment is merged into the corresponding path cluster; if the condition is not met, a new path cluster is constructed using the straight segment as the initial element.

[0136] Through the above processing, a cluster of straight paths is formed, consisting of multiple straight segments. Let the first... A cluster of straight paths is represented as:

[0137] ;

[0138] Furthermore, for straight path clusters Extract its endpoint candidate point set:

[0139] ;

[0140] Candidate point set of the endpoint Density-based spatial clustering is performed to obtain several endpoint subclusters:

[0141] ;

[0142] Among the endpoint subclusters, the two endpoint subclusters with the smallest and largest projected positions along the main direction of the path cluster are selected and respectively designated as straight path clusters. The trusted path start point and trusted path end point are denoted as, for example... Figure 3 :

[0143] ;

[0144] Similarly, after completing the incremental clustering and path endpoint identification of the straight sections, the same level of incremental clustering and endpoint extraction processing is performed on the set of curved sections.

[0145] Among the endpoint subclusters, the two endpoint subclusters with the smallest and largest projected positions along the main direction of the curve path cluster are selected and respectively designated as the curve path clusters. The trusted path start point and trusted path end point are denoted as:

[0146] .

[0147] Furthermore, the endpoint fusion of the trusted endpoint set of the path and the output path structure specifically includes:

[0148] By introducing a spatial location threshold, endpoints of straight path clusters and curved path clusters whose Euclidean distance is not greater than the threshold are merged into the same spatial node, forming a unified set of path boundary nodes.

[0149] Construct a straight sliding edge between the start and end nodes of the straight path cluster, and construct a turning sliding edge between the start and end nodes of the curved path cluster.

[0150] The airport taxiway network structure is composed of the set of nodes and the set of path edges.

[0151] In this embodiment, to eliminate minor deviations in spatial sampling between endpoints of different path clusters, a spatial location threshold is introduced. It is used to determine whether the endpoints represent the same taxiing network node.

[0152] The endpoints of straight path clusters and curved path clusters are considered to belong to the same spatial node when the following conditions are met:

[0153] ;

[0154] in: , , This represents the Euclidean distance. Endpoints satisfying the above conditions are merged to form unified path boundary nodes.

[0155] The fused set of trusted endpoints is denoted as:

[0156] ;

[0157] Each node This represents a spatial connection node in the airport taxiway network. Further: each cluster of straight paths... In its corresponding and Construct a straight sliding edge between them; each curve path cluster In its corresponding and Construct a turning and sliding edge between them.

[0158] This forms a set of nodes. With the set of path edges The airport taxiway network formed by these routes:

[0159] .

[0160] Please see Figure 4 An airport digital taxiway map modeling system based on flight trajectory data is used to implement the aforementioned airport digital taxiway map modeling method based on flight trajectory data. It includes a data preprocessing module 1, a trajectory segment identification module 2, a conflict segment processing module 3, a path clustering module 4, and a road network construction module 5. The trajectory segment identification module 2 is connected to the data preprocessing module 1, the conflict segment processing module 3 is connected to the trajectory segment identification module 2, the path clustering module 4 is connected to the conflict segment processing module 3, and the road network construction module 5 is connected to the path clustering module 4.

[0161] The data preprocessing module 1 is used to preprocess flight trajectories and establish a flight scene time-series trajectory subsequence dataset.

[0162] The trajectory segmentation and recognition module 2 is used to perform dual determination of vertical projection distance and angle projection on the flight scene time sequence subsequence dataset to identify potential straight-line segments and curve segments.

[0163] The conflict segment processing module 3 is used to adjust the time window for dividing the subsequence set, perform double judgment on the new subsequence dataset again, and process the shape boundary of the detected conflict segment to divide the determined straight line segment and curved segment.

[0164] The path clustering module 4 is used to perform incremental clustering on the straight line segment set and the curved line segment set respectively, and identify the reliable endpoint set of the path cluster.

[0165] The road network construction module 5 is used to perform endpoint fusion on the set of trusted endpoints of the path and output the airport taxiway network structure composed of nodes and path edges.

[0166] In this embodiment, after the airport digital taxiway map modeling system based on flight trajectory data is started, the data preprocessing module 1 first accesses the original positioning trajectory data of airport flights.

[0167] The data preprocessing module 1 groups the raw trajectory data according to the unique identifier of the aircraft and sorts them in ascending order by timestamp to form a time-series trajectory set for each aircraft.

[0168] Subsequently, the data preprocessing module 1, in conjunction with the preset polygon structure of the airport surface area, performs point-polygon inclusion detection on each trajectory point in the time-series trajectory set, filters out trajectory points located within the surface area, and extracts the surface taxiing trajectory sequence.

[0169] Finally, the data preprocessing module 1 divides the taxiing trajectory sequence into equal lengths by setting the starting position and length of the subsequence, constructs a flight scene time-series trajectory subsequence dataset, and transmits the constructed subsequence dataset to the trajectory segmentation and recognition module 2.

[0170] After receiving the subsequence dataset, the trajectory segmentation and recognition module 2 constructs a main direction vector from the starting point to the ending point for each subsequence, and calculates the vertical projection distance of each intermediate trajectory point in the subsequence relative to the main direction vector, as well as the angle between the direction vector of the intermediate trajectory point and the main direction vector.

[0171] The trajectory segmentation identification module 2 performs dual judgment based on the set vertical projection distance threshold and angle threshold: if all intermediate trajectory points in the subsequence satisfy that the vertical projection distance is not greater than the threshold and the included angle is not greater than the threshold, then the subsequence is judged as a straight segment and the internal trajectory points are marked as straight segment points; otherwise, it is judged as a curved segment and the internal trajectory points are marked as curved segment points. After completing the preliminary judgment, the trajectory segmentation identification module 2 sends the judgment result to the conflict segment processing module 3.

[0172] After receiving the preliminary judgment result, the conflict segment processing module 3 adjusts the starting position of the subsequence division and performs the dual judgment of vertical projection distance and angle projection on the new subsequence dataset again to obtain two configuration judgment results for each trajectory point.

[0173] The conflict segment processing module 3 determines the set of continuous trajectory points with inconsistent judgment results as a conflict segment, and divides the conflict segment into two parts based on the midpoint of the conflict segment. The first half is assigned to the straight segment set and the second half is assigned to the curve segment set, thereby completing the shape boundary processing of the conflict segment, dividing the determined straight segment and curve segment, and transmitting the processed straight segment set and curve segment set to the path clustering module 4.

[0174] After receiving the set of straight segments and the set of curved segments, the path clustering module 4 defines the start point, end point, and main direction angle for each straight segment, and the start point, end point, and curvature feature parameters for each curved segment. The path clustering module 4 performs incremental clustering on the straight segments according to the trajectory sequence, merging the straight segments that meet the conditions of direction consistency, spatial proximity, and main direction continuity into the same straight path cluster; the same method is used to perform incremental clustering on the set of curved segments to form a curved path cluster.

[0175] Subsequently, the path clustering module 4 extracts the endpoints of all segments within each path cluster, obtains endpoint subclusters through density-based spatial clustering, and selects the endpoint subclusters with the smallest and largest projection positions along the main direction of the path cluster as the trusted start and trusted end points of the path cluster, respectively, to generate a set of trusted endpoints for the path and transmit it to the road network construction module 5.

[0176] After receiving the set of reliable endpoints of the path, the road network construction module 5 introduces a spatial location threshold and merges the endpoints of the straight path cluster and the endpoints of the curved path cluster whose Euclidean distance is not greater than the threshold into the same spatial node, forming a unified set of path boundary nodes.

[0177] The road network construction module 5 constructs straight taxiing edges between the starting and ending nodes of each straight path cluster and constructs turning taxiing edges between the starting and ending nodes of each curved path cluster. Finally, the set of nodes and the set of path edges together constitute the airport taxiing road network structure and output it, thus completing the modeling of the airport digital taxiing map.

[0178] The above-disclosed embodiments are merely one or more preferred embodiments of this application and should not be construed as limiting the scope of this application. Those skilled in the art can understand that all or part of the processes for implementing the above embodiments and equivalent changes made in accordance with the claims of this application still fall within the scope of this application.

Claims

1. An airport digital taxiway map modeling method based on flight trajectory data, characterized in that, Includes the following steps: Flight trajectory data preprocessing is performed to establish a flight scene time-series trajectory subsequence dataset; By performing dual determination of vertical projection distance and angular projection on the flight scene time sequence subsequence dataset, potential straight-line segments and curve segments are identified; Adjust the time window for subsequence set division, perform double judgment again on the new subsequence dataset, and process the shape boundary of the detected conflict segments to divide the straight line segments and curved segments into defined segments. Incremental clustering is performed on the straight line segment set and the curved line segment set respectively to identify the credible endpoint set of the path cluster; Endpoint fusion is performed on the set of trusted endpoints of the path to output the airport taxiway network structure consisting of nodes and path edges.

2. The airport digital taxiway map modeling method based on flight trajectory data as described in claim 1, characterized in that, The process of preprocessing flight trajectories to establish a flight scene time-series trajectory subsequence dataset specifically includes: The original trajectory data is grouped and sorted based on the unique identifier of the aircraft to form a time-series trajectory set for each aircraft. By combining the airport surface area range, spatial filtering of trajectory points is performed to extract surface taxiing trajectory sequences; By setting the starting position and length of the subsequence, the scene gliding trajectory sequence is divided into equal-length segments to construct a time-sequence subsequence dataset.

3. The airport digital taxiway map modeling method based on flight trajectory data as described in claim 1, characterized in that, The dual determination of vertical projection distance and angular projection on the flight scene time-series subsequence dataset specifically includes: For each subsequence, construct the main direction vector based on its starting and ending points; Calculate the vertical projection distance of each intermediate trajectory point within the subsequence relative to the main direction vector, and the angle between the direction vector of that point and the main direction vector; Based on the set vertical projection distance threshold and angle threshold, the subsequence is determined to be a straight segment or a curved segment, and the internal trajectory point type is marked accordingly.

4. The airport digital taxiway map modeling method based on flight trajectory data as described in claim 1, characterized in that, The adjustment of the subsequence set into time windows and the shape boundary processing of conflict segments specifically include: By changing the starting position of the subsequence, the new subsequence dataset is subjected to double determination again, resulting in two configuration determination results for each trajectory point; The set of continuous trajectory points whose two judgment results are inconsistent is defined as the conflict segment; The conflict segment is divided into two parts, front and back, based on the midpoint of the conflict segment, and then assigned to the straight segment set and the curve segment set respectively, thus completing the stable structured segmentation.

5. The airport digital taxiway map modeling method based on flight trajectory data as described in claim 1, characterized in that, The incremental clustering of the straight segment set and the curved segment set, and the identification of the reliable endpoint set of the path cluster, specifically includes: Define geometric parameters for the set of straight segments and the set of curved segments, including the start point, end point, and direction or curvature characteristics; Incremental clustering is performed on straight segments according to the trajectory sequence order, and straight segments that meet the requirements of directional consistency, spatial proximity and main direction continuity are merged into straight path clusters; Extract the endpoints of all straight segments within the straight path cluster and perform density-based spatial clustering. Select the endpoint sub-clusters with the smallest and largest projections along the main direction as the reliable starting point and reliable ending point. The same method is applied to the set of curve segments to obtain the curve path cluster and its reliable endpoints.

6. The airport digital taxiway map modeling method based on flight trajectory data as described in claim 1, characterized in that, The endpoint fusion of the trusted endpoint set of the path, and the output path structure, specifically includes: By introducing a spatial location threshold, endpoints of straight path clusters and curved path clusters whose Euclidean distance is not greater than the threshold are merged into the same spatial node, forming a unified set of path boundary nodes. Construct a straight sliding edge between the start and end nodes of the straight path cluster, and construct a turning sliding edge between the start and end nodes of the curved path cluster. The airport taxiway network structure is composed of the set of nodes and the set of path edges.

7. The airport digital taxiway map modeling method based on flight trajectory data as described in claim 2, characterized in that, The unique identifier for an aircraft includes the aircraft number or flight number; The time-series trajectory set is obtained by sorting it in ascending order by timestamp; The airport surface area is a polygonal structure region, and the spatial filtering of trajectory points is achieved through point-polygon inclusion detection.

8. The airport digital taxiway map modeling method based on flight trajectory data as described in claim 3, characterized in that, The vertical projection distance is calculated by vector cross product and Euclidean norm; The included angle is calculated using the vector dot product and the inverse cosine function; If all intermediate trajectory points in a subsequence satisfy the condition that the vertical projection distance is not greater than the threshold and the included angle is not greater than the threshold, it is determined to be a straight segment; otherwise, it is determined to be a curved segment.

9. The airport digital taxiway map modeling method based on flight trajectory data as described in claim 5, characterized in that, The geometric parameters of the straight section include the start point, the end point, and the main direction angle; The geometric parameters of the curved section include the starting point, the ending point, and the curvature characteristic parameters; The density-based spatial clustering is implemented using the DBSCAN algorithm.

10. An airport digital taxiway map modeling system based on flight trajectory data, used to implement the airport digital taxiway map modeling method based on flight trajectory data as described in claim 1, characterized in that, It includes a data preprocessing module, a trajectory segmentation and identification module, a conflict segment processing module, a path clustering module, and a road network construction module. The trajectory segmentation and identification module is connected to the data preprocessing module, the conflict segment processing module is connected to the trajectory segmentation and identification module, the path clustering module is connected to the conflict segment processing module, and the road network construction module is connected to the path clustering module. The data preprocessing module is used to preprocess flight trajectories and establish a flight scene time-series trajectory subsequence dataset. The trajectory segmentation and recognition module is used to perform dual determination of vertical projection distance and angular projection on the flight scene time-series subsequence dataset to identify potential straight-line segments and curve segments. The conflict segment processing module is used to adjust the time window for dividing the subsequence set, perform double judgment again on the new subsequence dataset, and process the shape boundary of the detected conflict segment to divide the line segment and curve segment into a defined line segment and curve segment. The path clustering module is used to incrementally cluster the straight line segment set and the curved line segment set respectively, and identify the reliable endpoint set of the path cluster. The road network construction module is used to perform endpoint fusion on the set of trusted endpoints of the path and output the airport taxiway network structure composed of nodes and path edges.

Images