An airport aircraft taxiing efficiency review analysis method and system

Abstract

The application discloses an airport aircraft taxiing efficiency review analysis method and system, the method comprises the following steps: time-space alignment is carried out on flight plan, stand allocation, field monitoring radar and airport announcement data to generate a fusion time-space database; a weighted directed graph model with airport physical position as nodes and taxiable path as edges is constructed as an airport operation graph; abnormal taxiing time flights are identified, and actual trajectories are mapped to actual taxiing path sequences on the graph by using a hidden Markov map matching algorithm; the start and end nodes are used as constraints to calculate the theoretically optimal taxiing path sequence; the deviation nodes of the two paths are positioned, and the delay is attributed in combination with the node and edge attributes; finally, visual display and parameterized simulation deduction are carried out. The method and system realize automatic and accurate root cause tracing and strategy evaluation of the taxiing efficiency problem through data fusion and graph modeling technology, and improve the intelligent level of airport ground operation management.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of airport operation management, and in particular to an airport aircraft taxiing efficiency review analysis method and system. BACKGROUND

[0002] The airport ground operation efficiency is a key link affecting the punctuality rate and operation safety of flights. The ground taxiing time of an aircraft is affected by multi-dimensional dynamic factors such as stand allocation, taxiing route, runway usage mode, temporary construction, etc. Currently, the analysis of this problem has the following significant bottlenecks in the technical aspect:

[0003] Firstly, the technical obstacle of multi-source heterogeneous data fusion. Flight plans, stand allocation, high-precision radar trajectories, and real-time announcements are derived from independent systems, and their data formats, time and space references, and update frequencies are different, resulting in a serious "data island" phenomenon. There is a lack of effective technical means to perform high-precision time and space alignment, cleaning, and correlation on these data, making it difficult to build a unified time and space data base for in-depth analysis.

[0004] Secondly, the limitations of analysis models and calculation methods. Existing monitoring systems mostly focus on real-time status display and simple statistics, and lack a unified computable model that can represent the complex physical configuration and dynamic operation rules of the airport. Therefore, it is difficult to accurately restore and compare historical operation trajectories, and even more difficult to automatically calculate the theoretical optimal path and perform difference quantitative analysis through algorithms, resulting in root cause tracing remaining at the surface level of statistics and failing to locate specific spatial nodes and logical causes.

[0005] Thirdly, the lack of intelligent attribution and decision support tools. Current analysis results are mostly post-event reports, and data, models, and visual geographic information systems are not deeply integrated, especially lacking decision support capabilities based on simulation and deduction. Operation management personnel cannot quickly and intuitively assess the potential impact of different scheduling strategies on taxiing efficiency, and the decision-making process relies on experience, lacking quantitative, dynamic, and interactive technical assistance tools.

[0006] Therefore, the present field urgently needs to break through the existing technical framework and provide a technical solution that can systematically solve multi-source data fusion, build a computable analysis model, and realize automated attribution and strategy simulation, in order to improve the accuracy, depth, and intelligence level of airport ground operation analysis. SUMMARY

[0007] To solve the technical problems in the background art, the present application proposes an airport aircraft taxiing efficiency review analysis method and system.

[0008] The airport aircraft taxiing efficiency review analysis method proposed by the present application comprises the following steps:

[0009] S1. Perform spatiotemporal alignment processing on flight plan data, gate allocation data, airport surveillance radar data, and airport notice data from the same target airport to generate a fused spatiotemporal database;

[0010] S2. Based on the fused spatiotemporal database, construct a weighted directed graph model with airport physical locations as nodes, taxiable paths between nodes as edges, and edge weights dynamically determined by at least taxiing distance, average taxiing time based on historical taxiing data, and real-time operational constraints. This model serves as an airport operation map for taxiing path analysis and optimization.

[0011] S3. Based on the fused spatiotemporal database, identify flights with abnormal taxiing times, and map the actual trajectory data of the abnormal flights to the airport operation map using a map matching algorithm based on a hidden Markov model to generate an actual taxiing path sequence.

[0012] S4. In the airport operation map, using the start and end nodes of the actual taxiway sequence as constraints, calculate the corresponding theoretical optimal taxiway sequence through a graph search algorithm;

[0013] S5. Compare the actual taxiway path sequence with the theoretical optimal taxiway path sequence to locate the path deviation node, and perform delay attribution analysis based on the node attributes and edge attributes associated with the path deviation node in the airport operation map to obtain the delay attribution analysis results.

[0014] S6. Visualize and simulate the actual taxiing path sequence, the theoretical optimal taxiing path sequence, and the delay attribution analysis results.

[0015] Preferably, step S1 specifically includes:

[0016] Acquire heterogeneous operational data from flight schedule data sources, gate allocation data sources, airport surveillance radar data sources, and airport notice data sources;

[0017] All timestamps in heterogeneous operational data are uniformly converted to the standard Coordinated Universal Time (UTC) format.

[0018] The aircraft trajectory coordinates from the airport surveillance radar data source are uniformly projected to the airport's electronic data center using a coordinate transformation algorithm. Figure One A local plane coordinate system;

[0019] Based on aircraft call signs and time windows, the spatiotemporally aligned trajectory data is correlated and matched with the corresponding flight plans and gate allocation data;

[0020] The textual descriptions of construction areas in the airport notice data source are converted into polygon geofence data in the local plane coordinate system.

[0021] The completed associated and converted data is integrated with the pre-recorded airport basic geographic information data and structured operation constraint data, stored in a preset structured format, and a fusion space-time database is generated.

[0022] The airport basic geographic information data constitutes an airport configuration data basis, the structured operation constraint data and the airport announcement data constitute an operation constraint data basis, and the associated and matched flight plan, stand allocation and trajectory data constitute a historical taxiing analysis data basis.

[0023] Preferably, step S2 specifically comprises:

[0024] The airport configuration data basis and the operation constraint data basis are read from the fusion space-time database;

[0025] Based on the airport configuration data basis, the airport physical elements are abstracted as nodes of an airport operation graph, and the nodes at least include stand nodes, taxiway nodes, runway entrance nodes and construction area nodes;

[0026] Based on the operation constraint data basis, the physical connection between the nodes and the taxiing permission relationship are abstracted as edges of the airport operation graph, and the edges at least include taxiway segment edges and stand taxi-in / taxi-out edges;

[0027] A weight parameter is configured for each edge, and the weight parameter is at least determined by a taxiing distance, an average taxiing time based on the historical taxiing analysis data basis and a real-time operation constraint state, so as to complete the construction of the weighted directed graph model.

[0028] Preferably, step S3 specifically comprises:

[0029] From the historical taxiing analysis data basis of the fusion space-time database, the actual taxiing time of each flight in a target time period is extracted, combined with the aircraft type and the runway operation mode, and a statistical model based on a sliding window and a standard deviation threshold is used to identify flights with taxiing time exceeding an adaptive threshold, which are marked as taxiing time abnormal flights;

[0030] For each taxiing time abnormal flight, a continuous time-position point sequence is extracted from the corresponding field monitoring radar trajectory data as an actual trajectory;

[0031] The actual trajectory is taken as an observation sequence, the edges of the airport operation graph are taken as hidden states, a map matching algorithm based on a hidden Markov model is called, the most possible edge sequence corresponding to the actual trajectory is calculated, and the edge sequence is output as an actual taxiing path sequence.

[0032] Preferably, step S4 specifically comprises:

[0033] The start node and the end node of the actual taxiway path sequence are taken as the start node and the target node of the path search, respectively;

[0034] In the airport operation graph, based on the weight parameters of each edge, the edge sequence with the minimum cumulative weight from the start node to the target node is calculated by a graph search algorithm, and the edge sequence is output as the theoretical optimal taxiway path sequence.

[0035] Preferably, step S5 specifically comprises:

[0036] The edge sequence in the theoretical optimal taxiway path sequence is sequentially compared with the edge sequence in the actual taxiway path sequence, and edges that exist in the actual taxiway path sequence but do not exist in the theoretical optimal taxiway path sequence or have inconsistent sequences are extracted, and the nodes connected by the edges are defined as path deviation nodes;

[0037] In the airport operation graph, the nodes directly connected to the path deviation nodes and the associated edges are traversed, and the attribute information of the nodes and the attribute information of the associated edges are read and analyzed, respectively; wherein the attribute information of the nodes at least includes node type, construction state mark, associated gate number; the attribute information of the associated edges at least includes taxiway constraint condition, historical average waiting time;

[0038] Based on the path deviation nodes, the attribute information of the nodes and the attribute information of the associated edges, the cause of the deviation is determined to obtain a delay attribution analysis result containing the cause.

[0039] Preferably, the determination of the cause of the deviation specifically comprises:

[0040] If the path deviation node is a construction area node, or the attribute information of the associated edge indicates that it is in a construction closed state, it is determined that the cause is a construction area detour cause;

[0041] If the path deviation node is a gate node, and there is an edge in the actual taxiway path sequence that does not exist in the theoretical optimal taxiway path sequence and goes back and forth between different terminal areas or distant gate groups, it is determined that the cause is a gate allocation conflict cause;

[0042] If the historical average waiting time of the path deviation node associated edge exceeds a set threshold, it is determined that the cause is a taxiway conflict waiting cause.

[0043] Preferably, step S6 specifically comprises:

[0044] The visualization engine is called to load the two-dimensional or three-dimensional electronic map corresponding to the airport operation graph as a base map;

[0045] The actual taxiway path sequence is displayed in a first visual style and the theoretical optimal taxiway path sequence is displayed in a second visual style on the base map.

[0046] on the base map, according to the delay attribution analysis result, highlight mark the path deviation node and its related construction area, conflict point or non-optimal stand position in a third visual style;

[0047] provide an interactive interface to receive modification instructions for at least one operating parameter in the runway usage mode, construction area state or stand position allocation scheme;

[0048] based on the modified operating parameter, combine the airport operation graph and the discrete event simulation engine to deduce a new theoretical taxi path and estimated taxi time;

[0049] update the visual display content dynamically to reflect the deduction result, and generate an analysis report containing key indicator comparison and optimization suggestions.

[0050] Preferably, after step S6, it further comprises:

[0051] S7, based on the delay attribution analysis result and the simulation deduction result, generate a set of operation optimization instructions for the target airport; wherein the set of operation optimization instructions includes at least one of the following: stand re-allocation suggestion instruction, taxi route adjustment suggestion instruction, construction period optimization suggestion instruction.

[0052] The airport aircraft taxi efficiency review analysis system provided by the application comprises:

[0053] The multi-source spatio-temporal data fusion module is used for spatio-temporal alignment processing of flight plan data, stand allocation data, field monitoring radar data and airport announcement data from the same target airport, and generates a fusion spatio-temporal database;

[0054] The airport operation graph construction module is used for constructing a weighted directed graph model with airport physical positions as nodes, slidable paths between nodes as edges, and edge weight parameters determined at least by taxi distance, average taxi time based on historical taxi data statistics and real-time operation constraint state, as an airport operation graph for taxi path analysis and optimization;

[0055] The anomaly detection and processing module is used for identifying taxi time abnormal flights based on the fusion spatio-temporal database, and mapping the actual trajectory data of the abnormal flights to the airport operation graph through a map matching algorithm based on a hidden Markov model, to generate an actual taxi path sequence;

[0056] The data analysis module is used for calculating the corresponding theoretical optimal taxi path sequence by graph search algorithm in the airport operation graph with the start and end nodes of the actual taxi path sequence as constraints;

[0057] The data analysis module is also used for comparing the actual taxiing path sequence with the theoretical optimal taxiing path sequence to locate a path deviation node, and performing delay attribution analysis according to node attributes and edge attributes associated with the path deviation node in the airport operation graph to obtain a delay attribution analysis result.

[0058] The visual display and simulation deduction module is used for visual display and simulation deduction of the actual taxiing path sequence, the theoretical optimal taxiing path sequence and the delay attribution analysis result.

[0059] In the present application, the airport aircraft taxiing efficiency review analysis method and system proposed by the present application convert complex airport ground operation management into a calculable, traceable and simulative technical analysis process by constructing an airport operation graph model and a multi-source data fusion analysis framework, and overcome the limitations of relying on experience judgment and static rules. The present application realizes automatic and accurate positioning of aircraft taxiing efficiency abnormalities and multi-dimensional root cause tracing, improves the objectivity and efficiency of problem diagnosis, and provides intuitive and quantitative decision support for operation optimization through visual review and strategy simulation deduction, and enhances the collaboration and intelligence of the overall operation system. BRIEF DESCRIPTION OF DRAWINGS

[0060] Figure 1 A work flow diagram of an airport aircraft taxiing efficiency review analysis method proposed by the present application;

[0061] Figure 2 A system architecture diagram of an airport aircraft taxiing efficiency review analysis system proposed by the present application. DETAILED DESCRIPTION

[0062] REFERENCE Figure 1 and Figure 2 The airport aircraft taxiing efficiency review analysis method proposed by the present application includes the following steps:

[0063] S1, performing spatio-temporal alignment processing on flight plan data, gate allocation data, field monitoring radar data and airport announcement data from the same target airport to generate a fusion spatio-temporal database.

[0064] In the present embodiment, step S1 specifically includes: acquiring heterogeneous operation data from a flight plan data source, a gate allocation data source, a field monitoring radar data source and an airport announcement data source; uniformly converting all time stamps in the heterogeneous operation data into a standard coordinated universal time (UTC) time format; uniformly projecting aircraft trajectory coordinates in the field monitoring radar data source to an electronic airport chart through a coordinate conversion algorithm; and performing spatio-temporal alignment processing on the flight plan data, the gate allocation data, the field monitoring radar data and the airport announcement data from the same target airport to generate a fusion spatio-temporal database. Figure OneThe local plane coordinate system; based on the aircraft call sign and the time window, the trajectory data with completed space-time alignment is associated and matched with the corresponding flight plan and stand allocation data; the textual construction area description in the airport announcement data source is converted into polygon geographic fence data in the local plane coordinate system; the data after association and conversion is integrated with the pre-recorded airport basic geographic information data and structured operation constraint data, stored in accordance with the preset structured format, and a fusion space-time database is generated; wherein the airport basic geographic information data constitutes the airport configuration data basis, the structured operation constraint data and the airport announcement data constitute the operation constraint data basis, and the flight plan, stand allocation and trajectory data after association and matching constitute the historical taxiing analysis data basis.

[0065] It should be noted that the system accesses multiple operation data sources of the target airport through a data interface:

[0066] Flight plan and stand allocation data: obtained from the airport operation database (AODB), containing flight number, planned / actual take-off and landing time, allocated stand, aircraft type, etc.

[0067] Radar trajectory data of field monitoring: obtained from the multilateration system, recording the call sign, latitude, longitude, height, speed and time stamp accurate to milliseconds of the aircraft at a frequency of 1 per second.

[0068] Airport announcement data: obtained from the airport information management system, containing text information such as construction announcement and runway operation mode switching notice.

[0069] Specifically, the space-time alignment process is as follows:

[0070] Time alignment: convert the time stamps of all data sources into Coordinated Universal Time (UTC) to eliminate time deviations caused by time zone or system clock differences.

[0071] Spatial alignment: convert the WGS84 geographic coordinates of radar trajectories to the local independent coordinate system of the target airport through Gauss-Kruger projection algorithm. This coordinate system is consistent with the airport digital drawing reference, ensuring accurate comparison of all spatial element positions.

[0072] Based on the aircraft call sign and the accurate time window, the radar trajectory point sequence is automatically associated and matched with the corresponding flight record and stand allocation information to form a "flight stand continuous trajectory" data chain. At the same time, the text format construction announcement, such as "B taxiway middle section, 10:00-16:00 closed", is parsed and converted into a polygon geographic fence object in this local coordinate system and its effective time attribute.

[0073] The data that has been aligned, associated and converted is integrated with pre-recorded target airport basic geographic information data (including the precise geometry and topological relationship of all taxiways, stands and runways) and structured operation constraint data (such as the stand-aircraft matching table and fixed taxiway routing table), and is stored in a relational database or a spatio-temporal database according to a predefined Schema to form a unified fusion spatio-temporal database. The database provides a unique and consistent data source for all subsequent analysis steps.

[0074] S2, based on the fusion spatio-temporal database, a weighted directed graph model is constructed, which takes the airport physical location as a node, takes the slidable path between nodes as an edge, and the weight parameter of the edge is at least determined by the taxi distance, the average taxi time based on historical taxi data statistics and the real-time operation constraint state, as an airport operation graph for taxi path analysis and optimization.

[0075] In the embodiment, step S2 specifically comprises:

[0076] The airport configuration data basis and the operation constraint data basis are read from the fusion spatio-temporal database.

[0077] Based on the airport configuration data basis, the airport physical elements are abstracted as nodes of the airport operation graph, and the nodes at least include stand nodes, taxiway nodes, runway entrance nodes and construction area nodes.

[0078] Based on the operation constraint data basis, the physical connection and the taxi permission relationship between the nodes are abstracted as edges of the airport operation graph, and the edges at least include taxiway segment edges and stand taxi-in and taxi-out edges.

[0079] A weight parameter is configured for each edge, and the weight parameter is at least determined by the taxi distance, the average taxi time based on the historical taxi analysis data basis statistics and the real-time operation constraint state, so as to complete the construction of the weighted directed graph model.

[0080] Specifically, to complete the construction of the weighted directed graph model, the system configures a comprehensive and dynamically updated weight parameter for each edge in the airport operation graph, which directly determines the passage cost calculation in the subsequent path search algorithm. The configuration of the weight parameter is not static assignment, but is dynamically generated through the following technical process:

[0081] The taxi distance cost component: based on the precise coordinates stored in the airport basic geographic information data, the geometric length (unit: meter) of the physical taxi path represented by each edge is calculated. The length is divided by a reference taxi speed (for example, a typical taxi speed of 25 knots obtained according to the airport regulations or historical data statistics), and is converted into a basic time cost This part reflects the basic time required to pass through the segment under ideal conditions without obstacles and conflicts.

[0082] Historical average taxi time component: The system periodically extracts data from the historical taxi analysis data base. For each edge, the system queries the records of all actual taxi trajectories passing through the edge in the recent period (e.g. past 24 hours or past one week), and calculates the average additional waiting or delay time of these flights on the edge (or its associated node) . This statistic can be further divided by different time periods (e.g. peak / off-peak), different runway operation modes, to obtain more representative historical cost. This component quantifies the typical time loss of the road segment due to regular traffic patterns, structural bottlenecks, etc.

[0083] Real-time operation constraint state penalty component: The system monitors the real-time operation constraint data base and integrates dynamic information in the spatio-temporal database. If a node or area associated with an edge is marked as a specific constraint state at the current time (e.g. the associated construction area node is in the “active” state, causing the edge to be closed or limited; or according to real-time traffic management, the edge is subjected to flow control), a significant, numerical time penalty value is applied to the edge . This penalty value can be pre-defined or dynamically calculated according to the severity of the constraint, such as a construction closure that may result in a very long detour, and the penalty value can be set to hundreds or even thousands of seconds.

[0084] Finally, the comprehensive weight W of each edge at time t is calculated by a weighted summation model of the above components. The calculation formula is specifically:

[0085] ;

[0086] wherein, are the harmonic coefficients of the corresponding components, used to balance the influence degree of different components, and their values can be calibrated according to the characteristics of airport operation; is the average additional waiting or delay time; is the basic time cost; is the time penalty value.

[0087] It should be noted that the system recalculates and updates the weights of all edges at a set period (e.g. every minute) when constructing the graph, so as to ensure that the weighted directed graph model can dynamically and truly reflect the real-time traffic cost of the airport ground.

[0088] S3, based on the integrated spatio-temporal database, identify taxi time abnormal flights, and map the actual trajectory data of abnormal flights to the airport operation graph through a map matching algorithm based on a hidden Markov model, to generate an actual taxi path sequence.

[0089] In this embodiment, step S3 specifically comprises: extracting actual taxiing time of each flight in the target time period from the historical taxiing analysis data base of the spatio-temporal database, combining the aircraft model and the runway operation mode, and using a statistical model based on a sliding window and a standard deviation threshold to identify flights with taxiing time exceeding the adaptive threshold, and marking the flights as taxiing time abnormal flights; for each taxiing time abnormal flight, extracting a continuous time-position point sequence as an actual trajectory from the corresponding field monitoring radar trajectory data; taking the actual trajectory as an observation sequence and taking the edges of the airport operation graph as hidden states, calling a map matching algorithm based on a hidden Markov model, calculating the most possible edge sequence corresponding to the actual trajectory, and outputting the edge sequence as an actual taxiing path sequence.

[0090] Specifically, the abnormal flight identification process comprises:

[0091] From the historical taxiing analysis data base of the spatio-temporal database, the measured taxiing time of all inbound / outbound flights in the past 24 hours in the "runway to stand" or "stand to runway" stage is extracted. For the same aircraft model and the same runway operation mode, the sliding window statistical method is used to calculate the mean value and the standard deviation of the taxiing time. Flights with taxiing time exceeding are automatically marked as taxiing time abnormal flights.

[0092] Specifically, the trajectory map matching process comprises:

[0093] For the flight CA1234 marked as abnormal, a dense trajectory point sequence is extracted from the associated radar data. A map matching algorithm based on a hidden Markov model (HMM) is used to match the continuous trajectory point sequence to the edges of the discrete airport operation graph. Specifically, each trajectory point is taken as an observation value and the edges of the graph are taken as hidden states. By calculating the probability of the observation points falling on each edge (emission probability) and the possibility of the transition between edges (transition probability), the most possible edge sequence is solved by using the Viterbi algorithm. The edge sequence is the actual taxiing path sequence, for example, [E101, E205, E308, E412].

[0094] S4, in the airport operation graph, the start and end nodes of the actual taxiing path sequence are taken as constraints, and the corresponding theoretical optimal taxiing path sequence is calculated by a graph search algorithm.

[0095] In this embodiment, step S4 specifically comprises: taking the start node and the end node of the actual taxiing path sequence as the start node and the target node of the path search, respectively; in the airport operation graph, based on the weight parameters of each edge, the edge sequence with the minimum cumulative weight from the start node to the target node is calculated by a graph search algorithm, and the edge sequence is output as the theoretical optimal taxiing path sequence.

[0096] Specifically, the start and end nodes of the obtained actual taxiing path sequence are taken as the start point and target of path planning. On the constructed airport operation graph, the dynamic weight of each edge is taken as the passing cost, and A Shortest path search algorithm. The algorithm combines the actual traveled cost (cumulative weight from the start point to the current node) and the heuristic estimated cost (straight-line distance from the current node to the target estimated time), and efficiently searches for the edge sequence with the minimum cumulative weight from the start point to the target, i.e., the theoretical optimal taxiing path sequence, such as [E101, E203, E310, E412].

[0097] S5, compare the actual taxiing path sequence with the theoretical optimal taxiing path sequence to locate the path deviation node, and perform delay attribution analysis according to the node attributes and edge attributes associated with the path deviation node in the airport operation graph to obtain a delay attribution analysis result.

[0098] Specifically, the delay attribution analysis result includes at least one of the following information: cause classification leading to the delay, identification of the path deviation node, and quantitative contribution evaluation of each cause to the total delay time.

[0099] In this embodiment, step S5 specifically includes: sequentially comparing the edge sequence in the theoretical optimal taxiing path sequence with the edge sequence in the actual taxiing path sequence, extracting edges that exist in the actual taxiing path sequence but do not exist in the theoretical optimal taxiing path sequence or edges that are inconsistent in sequence, and defining the nodes connected by these edges as path deviation nodes; in the airport operation graph, traversing the nodes directly connected to the path deviation nodes and their associated edges, respectively reading and analyzing the attribute information of the nodes and the attribute information of the associated edges; wherein the attribute information of the nodes at least includes node type, construction state marker, and associated gate number; the attribute information of the associated edges at least includes taxiing constraint condition and historical average waiting time; based on the path deviation node, the attribute information of the nodes, and the attribute information of the associated edges, the cause of the deviation is determined to obtain a delay attribution analysis result containing the cause.

[0100] Specifically, determining the cause of the deviation specifically includes:

[0101] If the path deviation node is a construction area node, or the attribute information of the associated edge indicates that it is in a construction closed state, it is determined that the cause is a construction area detour cause;

[0102] If the path deviation node is a gate node, and there is an edge in the actual taxiing path sequence that does not exist in the theoretical optimal taxiing path sequence and goes back and forth between different terminal areas or distant gate groups, it is determined that the cause is a gate allocation conflict cause;

[0103] If the historical average waiting time of the path deviation node associated edge exceeds the set threshold, the cause is determined to be a taxi conflict waiting cause.

[0104] Specifically, the actual path sequence [E101, E205, E308, E412] is compared with the theoretical optimal path sequence [E101, E203, E310, E412] edge by edge. It is found that the actual path selects E205 instead of E203 after the second node, resulting in completely different subsequent paths. Node Nx (the starting node of E205) is located as the key path deviation node.

[0105] Further, in the airport operation graph, the attributes of the deviation node Nx and its directly associated nodes and edges are queried. It is found that the node Nx is associated with a construction announcement edge pointing to the construction node CZ1, and the state of CZ1 is “active” during the CA1234 taxi period. The attributes of edge E203 are queried, and its “traffic state” is marked as “closed” during the same period. The attributes of edge E205 are queried, and its “historical average waiting time” is displayed as 120 seconds.

[0106] Specifically, the system automatically concludes according to the preset causal inference logic:

[0107] The main cause (construction detour): due to the closure of E203 in the optimal path due to construction, the aircraft is forced to turn to E205 at Nx.

[0108] The secondary cause (node congestion): the detour path E205 itself has a high historical waiting time.

[0109] The system generates a structured delay cause analysis result, including cause classification, associated node / edge identification, and estimates that construction detour causes about 180 seconds of additional taxi, node congestion causes about 120 seconds of waiting, and quantifies the contribution degree of each factor.

[0110] S6, visualizing and simulating the actual taxi path sequence, the theoretical optimal taxi path sequence, and the delay cause analysis result.

[0111] In the embodiment, step S6 specifically comprises: calling a visualization engine, loading a two-dimensional or three-dimensional electronic map corresponding to the airport operation graph as a base map; superimposedly displaying the actual taxi path sequence in a first visual style and the theoretical optimal taxi path sequence in a second visual style on the base map; according to the delay attribution analysis result, highlighting and marking the path deviation node and its related construction area, conflict point or non-optimal stand in a third visual style on the base map; providing an interactive interface to receive a modification instruction for at least one of the runway usage mode, the construction area state or the stand allocation scheme; based on the modified operation parameter, combining the airport operation graph and the discrete event simulation engine to deduce a new theoretical taxi path and an estimated taxi time; dynamically updating the visualization display content to reflect the deduction result and generating an analysis report containing a comparison of key indicators and optimization suggestions.

[0112] Specifically, the visualization display process comprises: the system calling a WebGL engine, drawing the actual taxi path with a red highlight line, the theoretical optimal path with a green dashed line, and the deviation node Nx and the construction area CZ1 with a flashing icon on the three-dimensional airport electronic map. The side information panel lists the taxi time comparison of CA1234 flight, the attribution analysis result and the quantitative data in detail. Among them, the user can modify the operation parameters in the interface. For example, manually set the state of the construction node CZ1 from "activated" to "inactivated". The system immediately re-runs the path search and time estimation based on the modified parameters (at this time, the E203 edge weight returns to normal). The simulation result shows that if the construction does not occur, CA1234 adopts the optimal path, and the estimated taxi time can be reduced by about 300 seconds. The interface is dynamically updated, and an analysis report is automatically generated, summarizing the root cause of the problem.

[0113] In the embodiment, it also comprises:

[0114] S7, based on the delay attribution analysis result and the simulation deduction result, generating a set of operation optimization instructions for the target airport; wherein the operation optimization instruction set comprises at least one of the following: stand re-allocation suggestion instruction, taxi route adjustment suggestion instruction, construction period optimization suggestion instruction.

[0115] With reference to Figure 1 and Figure 2 , the present application proposes an airport aircraft taxi efficiency review analysis system, comprising:

[0116] A multi-source spatio-temporal data fusion module is configured to perform spatio-temporal alignment processing on flight plan data, stand allocation data, field monitoring radar data and airport announcement data from the same target airport to generate a fused spatio-temporal database;

[0117] An airport operation graph construction module is configured to construct a weighted directed graph model with airport physical locations as nodes, slidable paths between nodes as edges, and edge weight parameters determined at least by slidable distances, average slidable times based on historical slidable data statistics, and real-time operation constraint states, as an airport operation graph for slidable path analysis and optimization based on the fusion spatio-temporal database;

[0118] An abnormality detection and processing module is configured to identify abnormal flight based on the fusion spatio-temporal database, and map actual trajectory data of the abnormal flight to the airport operation graph through a map matching algorithm based on a hidden Markov model to generate an actual slidable path sequence.

[0119] A data analysis module is configured to calculate a corresponding theoretical optimal slidable path sequence by a graph search algorithm with start and end nodes of the actual slidable path sequence as constraints in the airport operation graph.

[0120] The data analysis module is further configured to compare the actual slidable path sequence with the theoretical optimal slidable path sequence to locate path deviation nodes, and perform delay attribution analysis according to node attributes and edge attributes associated with the path deviation nodes in the airport operation graph to obtain a delay attribution analysis result.

[0121] A visual display and simulation deduction module is configured to visually display and simulate deduction of the actual slidable path sequence, the theoretical optimal slidable path sequence, and the delay attribution analysis result.

[0122] The above is only a preferred embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can make equivalent replacements or changes to the technical solution and inventive concept of the present application within the technical scope disclosed by the present application, which should be covered by the protection scope of the present application.

Claims

1. A method for retrospective analysis of airport aircraft taxiing efficiency, characterized in that, Includes the following steps: S1. Perform spatiotemporal alignment processing on flight plan data, gate allocation data, airport surveillance radar data, and airport notice data from the same target airport to generate a fused spatiotemporal database; S2. Based on the fused spatiotemporal database, construct a weighted directed graph model with airport physical locations as nodes, taxiable paths between nodes as edges, and edge weights dynamically determined by at least taxiing distance, average taxiing time based on historical taxiing data, and real-time operational constraints. This model serves as an airport operation map for taxiing path analysis and optimization. S3. Based on the fused spatiotemporal database, identify flights with abnormal taxiing times, and map the actual trajectory data of the abnormal flights to the airport operation map using a map matching algorithm based on a hidden Markov model to generate actual taxiing path sequences. S4. In the airport operation map, using the start and end nodes of the actual taxiway sequence as constraints, calculate the corresponding theoretical optimal taxiway sequence through a graph search algorithm; S5. Compare the actual taxiway path sequence with the theoretical optimal taxiway path sequence to locate the path deviation node, and perform delay attribution analysis based on the node attributes and edge attributes associated with the path deviation node in the airport operation map to obtain the delay attribution analysis results. S6. Visualize and simulate the actual taxiing path sequence, the theoretical optimal taxiing path sequence, and the delay attribution analysis results.

2. The airport aircraft taxiing efficiency retrospective analysis method according to claim 1, characterized in that, Step S1 specifically includes: Acquire heterogeneous operational data from flight schedule data sources, gate allocation data sources, airport surveillance radar data sources, and airport notice data sources; All timestamps in heterogeneous operational data are uniformly converted to the standard Coordinated Universal Time (UTC) format. The aircraft trajectory coordinates from the airport surveillance radar data source are uniformly projected to the local plane coordinate system consistent with the airport electronic map through a coordinate transformation algorithm. Based on aircraft call signs and time windows, the spatiotemporally aligned trajectory data is correlated and matched with the corresponding flight plans and gate allocation data; The textual descriptions of construction areas in the airport notice data source are converted into polygon geofence data in the local plane coordinate system. The data after association and transformation is integrated with the pre-entered airport basic geographic information data and structured operational constraint data, stored in a preset structured format, and a fused spatiotemporal database is generated. Among them, airport basic geographic information data constitutes the basis of airport configuration data, structured operation constraint data and airport notice data constitute the basis of operation constraint data, and the associated and matched flight plans, gate allocation and trajectory data constitute the basis of historical taxiing analysis data.

3. The airport aircraft taxiing efficiency retrospective analysis method according to claim 1 or 2, characterized in that, Step S2 specifically includes: Read the airport configuration data base and operational constraint data base from the fused spatiotemporal database; Based on airport configuration data, airport physical elements are abstracted into nodes of the airport operation map. These nodes include at least aircraft stand nodes, taxiway nodes, runway entrance nodes, and construction area nodes. Based on operational constraint data, the physical connections and taxiing permission relationships between nodes are abstracted into edges of the airport operation map, which include at least taxiway segment edges and aircraft stand taxiing in and taxiing out edges; Each edge is configured with a weight parameter, which is determined by at least the gliding distance, the average gliding time based on historical gliding analysis data, and the real-time running constraint state, thereby completing the construction of the weighted directed graph model.

4. The airport aircraft taxiing efficiency retrospective analysis method according to claim 1 or 2, characterized in that, Step S3 specifically includes: Based on the historical taxiing analysis data of the integrated spatiotemporal database, the actual taxiing time of each flight within the target time period is extracted. Combined with the aircraft type and runway operation mode, a statistical model based on sliding window and standard deviation threshold is used to identify flights whose taxiing time exceeds the adaptive threshold and mark them as abnormal taxiing time flights. For each flight with abnormal taxiing time, a continuous time-location sequence is extracted from its corresponding airport monitoring radar trajectory data as the actual trajectory. The actual trajectory is used as the observation sequence, and the edges of the airport operation map are used as the hidden state. A map matching algorithm based on the Hidden Markov Model is called to calculate the edge sequence corresponding to the actual trajectory, and the edge sequence is output as the actual taxiing path sequence.

5. The airport aircraft taxiing efficiency retrospective analysis method according to claim 1, characterized in that, Step S4 specifically includes: The starting and ending nodes of the actual gliding path sequence are used as the starting and target nodes of the path search, respectively. In the airport operation map, based on the weight parameters of each edge, the graph search algorithm is used to calculate the edge sequence with the minimum cumulative weight from the starting node to the target node, and the edge sequence is output as the theoretically optimal taxiing path sequence.

6. The airport aircraft taxiing efficiency retrospective analysis method according to claim 1, characterized in that, Step S5 specifically includes: The edge sequence in the theoretical optimal gliding path sequence is compared with the edge sequence in the actual gliding path sequence in order. Edges that exist in the actual gliding path sequence but not in the theoretical optimal gliding path sequence, or edges with inconsistent order, are extracted, and the nodes connected by these edges are defined as path deviation nodes. In the airport operation map, the nodes directly connected to the path deviation node and their associated edges are traversed, and the attribute information of the nodes and the attribute information of the associated edges are read and analyzed respectively; wherein, the attribute information of the nodes includes at least node type, construction status mark, and associated aircraft stand number; the attribute information of the associated edges includes at least taxiing constraints and historical average waiting time. Based on the path deviation nodes, node attribute information, and associated edge attribute information, the causes of deviation are determined to obtain delay attribution analysis results that include the causes.

7. The airport aircraft taxiing efficiency retrospective analysis method according to claim 6, characterized in that, The determination of the causes of the deviation specifically includes: If the path deviation node is a construction area node, or the attribute information of its associated edge indicates that it is in a construction closed state, then the cause is determined to be a detour due to the construction area. If the path deviation node is a gate node, and there are edges in the actual taxiing path sequence that do not exist in the theoretical optimal taxiing path sequence, and that travel between different terminal areas or distant gate groups, then the cause is determined to be a gate allocation conflict. If the historical average waiting time of the associated edge of the path deviation node exceeds the set threshold, the cause is determined to be a sliding conflict waiting cause.

8. The airport aircraft taxiing efficiency retrospective analysis method according to claim 1, characterized in that, Step S6 specifically includes: Call the visualization engine and load the two-dimensional or three-dimensional electronic map corresponding to the airport operation map as the base map; On the base map, the actual gliding path sequence is overlaid in a first visual style, and the theoretically optimal gliding path sequence is overlaid in a second visual style. On the base map, based on the delay attribution analysis results, path deviation nodes and their related construction areas, conflict points or suboptimal camera positions are highlighted in a third-view style. Provides an interactive interface to receive modification instructions for at least one of the following operating parameters: runway usage mode, construction area status, or aircraft stand allocation scheme; Based on the modified operating parameters, combined with the airport operation map and discrete event simulation engine, a new theoretical taxiing path and estimated taxiing time are derived. The visualization content is dynamically updated to reflect the simulation results, and an analysis report containing key indicator comparisons and optimization suggestions is generated.

9. The airport aircraft taxiing efficiency retrospective analysis method according to claim 1, characterized in that, Following step S6, the following is also included: S7. Based on the delay attribution analysis results and simulation results, generate an operation optimization instruction set for the target airport; wherein, the operation optimization instruction set includes at least one of the following: stand reallocation suggestion instruction, taxiway route adjustment suggestion instruction, and construction period optimization suggestion instruction.

10. A system for reviewing and analyzing the taxiing efficiency of airport aircraft, characterized in that, include: The multi-source spatiotemporal data fusion module is used to perform spatiotemporal alignment processing on flight plan data, gate allocation data, airport surveillance radar data and airport notice data from the same target airport to generate a fused spatiotemporal database. The airport operation map construction module is used to construct a weighted directed graph model based on the fused spatiotemporal database. The model has airport physical locations as nodes, taxiable paths between nodes as edges, and the weight parameters of the edges are dynamically determined by at least the taxi distance, the average taxi time based on historical taxi data, and the real-time operation constraint status. This model serves as an airport operation map for taxi path analysis and optimization. An anomaly detection and processing module is used to identify flights with abnormal taxiing times based on the fused spatiotemporal database, and to map the actual trajectory data of the abnormal flights to the airport operation map through a map matching algorithm based on a hidden Markov model, thereby generating an actual taxiing path sequence. The data analysis module is used to calculate the theoretically optimal taxiway sequence in the airport operations map by using a graph search algorithm, with the start and end nodes of the actual taxiway sequence as constraints. The data analysis module is also used to compare the actual taxiway path sequence with the theoretical optimal taxiway path sequence to locate the path deviation node, and to perform delay attribution analysis based on the node attributes and edge attributes associated with the path deviation node in the airport operation map to obtain the delay attribution analysis results. The visualization and simulation module is used to visualize and simulate the actual taxiing path sequence, the theoretical optimal taxiing path sequence, and the delay attribution analysis results.

Images