Multi-source data fused sorting method and system for airport flight coordinated launch

Abstract

The invention belongs to the technical field of air transportation management, and discloses a multi-source data fused sorting method and system for airport flight coordinated launch. The method comprises the following steps: collecting a multi-source data source of a flight; performing multi-source data fusion, generating a five-dimensional dynamic situation matrix, and reflecting the comprehensive state of the flight in real time; judging whether manual priority ranking processing is carried out or not; by constructing an intelligent decision-making system for aircraft departure sorting, optimization target decision-making and intelligent sorting are carried out; the visual sorting interface displays the alternative push-out sorting schemes; obtaining a sorting result of flight coordination deduction; establishing a dynamic reentry mechanism for flights with updated resources or changed states, and reinserting the flights into the global sorting queue; and balancing the adaptation degree of the original priority and the new state of the flight through a dynamic weight distribution algorithm. The airport flight scheduling efficiency is improved, so that flight delay is reduced, the operation efficiency is improved, and the passenger satisfaction degree is enhanced.

Description

Technical Field

[0001] This invention belongs to the field of air transport management technology, and in particular relates to a sorting method and system for airport flight coordination that integrates multi-source data. Background Technology

[0002] Traditional departure sequencing uses a release sequencing mechanism based on fixed rules, but it faces two challenges in actual operation: On the one hand, the dynamic uncertainty of ground support (such as delays in catering and cargo and mail loading and unloading) causes the preset sequencing scheme to become disconnected from the actual operating status, directly affecting the accuracy of flight departure times; on the other hand, the limited availability of trailer resources and insufficient scheduling efficiency lead to resource contention, and crews often occupy trailer resources in advance in order to obtain timely pushback instructions, resulting in a vicious cycle of trailer idleness and time window squeezing.

[0003] Traditional dispatching systems suffer from insufficient data integration capabilities, relying mainly on localized structured data such as flight schedules and some support nodes. This results in key parameters such as weather evolution, airspace traffic dynamics, and passenger gathering behavior not being incorporated into the decision-making model in real time.

[0004] Traditional static rules and offline optimization models have drawbacks such as high decision-making delays and high risk of secondary delays when dealing with disturbances such as temporary airspace control and sudden aircraft failures.

[0005] Traditional models focus on a single objective, such as minimizing delay time, and often neglect the balance of indicators such as safe intervals, fuel consumption, and passenger satisfaction. Summary of the Invention

[0006] To overcome the problems existing in related technologies, the present invention discloses an embodiment of a sorting method and system for airport flight coordination that integrates multi-source data.

[0007] The technical solution is as follows: a sorting method derived from airport flight coordination by integrating multi-source data, including the following steps: S1 collects data from multiple data sources for flights; these multiple data sources include: flight data sources, ground support data sources, meteorological data sources, passenger data sources, and air traffic control data sources. S2 integrates the collected data from multiple sources to generate a five-dimensional dynamic situation matrix, which reflects the overall status of flights in real time. S3. Based on the obtained comprehensive status, determine whether to perform manual priority sorting; if yes, proceed to step S4; otherwise, proceed to step S5. S4, based on the results of manual sorting and priority processing, optimizes target decision-making and intelligent sorting by constructing an intelligent decision-making system for aircraft departure sorting; S5, based on the real-time flight sequence optimization results, performs multi-dimensional visual early warnings on real-time flights with flight bars that include adjacent gate prompts and manual intervention functions, and displays the early warning results in a visual sorting interface to present alternative sorting schemes. S6: For the obtained alternative launch sorting schemes, perform critical coordination pool processing, real-time support progress monitoring, trailer resource allocation optimization, and delay cause analysis to obtain the flight coordination launch sorting results; S7: Based on the sorting results obtained from flight coordination, a dynamic re-entry mechanism is established for flights with resource updates or status changes, and the flights are re-inserted into the global sorting queue; through a dynamic weight allocation algorithm, the original priority of the flights is balanced with the adaptability of the new status.

[0008] In step S2, a five-dimensional dynamic situation matrix is ​​generated, including: A batch processing framework is constructed, and a quaternion interpolation algorithm is used to align the spatiotemporal reference of multi-source data to generate a five-dimensional dynamic situation matrix. The expression is: ; In the formula, For flight data, To provide ground support for data, For meteorological data, For passenger data, This is air traffic control data.

[0009] In step S3, based on the obtained overall status, it is determined whether to perform manual priority sorting, including: The improved two-way cognitive collaboration mechanism is achieved by manually adjusting flight sequences through a drag-and-drop interactive interface and manual input of sorting priorities. The improved two-way cognitive collaboration mechanism includes: actively identifying manual adjustment intentions and recommending the optimal adjustment range; triggering a priority overriding mechanism when a manual intervention signal is detected; verifying operation permissions through an enhanced multi-factor authentication framework; and writing the adjusted sequence into a distributed cache queue with version control capabilities. The constraint propagation method is used to dynamically correct the constraints of subsequent flights: a flight sequence constraint relationship network is constructed, the impact of manual adjustments on related flights is calculated, the minimum adjustment principle is applied to only correct the constraints of necessary flights, and the sorting position of all affected flights is optimized through recursive update technology.

[0010] In step S4, by constructing an intelligent decision-making system for aircraft departure sequencing, optimization target decisions and intelligent sequencing are performed, including: By coupling multi-dimensional parameters and adjusting dynamic weights, an intelligent decision-making system for aircraft departure sequencing is constructed. A comprehensive scoring function is established, and multiple key factors are considered through weighted summation to determine the priority of flight sequencing. The expression is as follows: ; In the formula, As a weighting factor for delay costs, For safety cost weighting coefficient, This is a weighting coefficient for the airport's on-time departure rate. This is a weighting coefficient for the airport's on-time departure rate. For passenger experience parameters, weighting coefficients For airlines, the weighting coefficient is... To address delay costs, a delay cost matrix is ​​constructed, categorized by aircraft type and airline. Statutory compensation standards and carbon emission conversion cost indicators are introduced, and the cost coefficient is dynamically adjusted by training the model using airline operational data. For safety and cost reasons, the standard computer type combination interval is affected by time slot limitations; The airport on-time departure rate parameter is defined as a parameter affecting the airport's overall departure rate. For passenger experience parameters, passenger impact parameters; For airline weight, market share, historical on-time rate, historical departure status, VIP flight marking; Substituting the real-time status data of each flight into the above formula, a comprehensive score for each flight is calculated. The priority ranking of flights is determined based on the scores, forming a baseline intelligent launch ranking scheme. Based on the baseline ranking scheme, real-time adjustments are made when the following situations occur: airlines temporarily requesting priority adjustments, VIP flights urgently jumping the queue, sudden equipment failures affecting specific flights, temporary changes in air traffic control instructions, and sudden changes in weather conditions. The system recalculates the priority of affected flights based on temporary changes and dynamically adjusts the ranking sequence to ensure that the ranking scheme always adapts to the actual operational needs on site.

[0011] In step S5, the adjacent aircraft stand warning function uses a topology space modeling and dynamic risk assessment system to predict and warn of potential taxiing conflicts between aircraft stands. The use of a topology space modeling and dynamic risk assessment system to predict and warn of potential taxiing conflicts between aircraft stands includes: S5.1 Identification of Adjacent Camera Positions: System construction station topology diagram Among them, vertex set Represents the machine station node and edge set. Represents the intersection relationships of potential gliding paths; each edge Carrying weight vector , These represent physical distance, gliding angle difference, path intersection count, and historical conflict frequency, respectively; the historical conflict frequency is dynamically updated through gliding trajectory data to achieve adaptive optimization of topology relationships; Construct a conflict risk assessment model and define the aircraft position pair risk factor : ; In the formula, All are dynamically weighted coefficients. This is a spatial correlation function. This is a time window overlap function. For aircraft characteristic functions, For meteorological influence functions, For the camera position The weight vector, For camera position and camera position At any moment The overlap of the launch time windows For camera position At any moment Aircraft characteristic parameters, For camera position At any moment Aircraft characteristic parameters, For a moment Meteorological state parameters; Using a sliding time window, the risk of gliding operations within the next T minutes is predicted, forming a spatiotemporal risk heat matrix. When the risk coefficient exceeds the threshold, an early warning mechanism is triggered. S5.2 Visual Early Warning; Based on the calculated risk coefficient To achieve multi-dimensional visual early warning, including: Color coding maps the risk level to the HSV color space, with low risk displayed as green, medium risk as yellow, high risk as orange, and critical risk as red with an added flashing effect; Interactive topology map, which overlays dynamic topology relationship layers on flight bars, uses line thickness, color, and solidity to represent the strength of association and risk of conflict; 3D scene reconstruction: For complex aircraft stand layouts, it provides 3D scene reconstruction function, overlays aircraft movement trajectory and conflict hotspot areas to obtain potential conflict causes and impact range; Simultaneously, based on the detected conflict risks, alternative pushback sequencing schemes are automatically generated. After detecting taxiing conflict risks between aircraft stands, the sequencing formula is automatically recalculated to generate multiple sequencing combinations that avoid conflicts. Each scheme adjusts the pushback sequence of different flights to ensure that taxiing paths do not intersect or conflict. The risk changes brought about by the scheme adjustments are displayed in the form of a difference heatmap, comparing the risk coefficients of the original scheme and the alternative schemes, and using color depth to indicate the degree of risk change: green indicates reduced risk, red indicates increased risk, and the darker the color, the greater the change, intuitively showing the risk impact of each adjustment scheme. In conjunction with the trailer resource allocation and critical coordination pool handling in step S6, the overall conflict risk is minimized. The visual sorting interface displays a countdown to the estimated departure time and restricted time based on the number of flights and control positions, with color changes indicating that ground support and apron control should push back and take off within the estimated departure time or restricted time.

[0012] In step S6, the critical coordination pool processing includes: employing a dynamic triggering mechanism based on multimodal data fusion, and constructing a priority evaluation model based on a time window decay function by real-time collection of flight support progress, air traffic control instructions, and resource availability status. This priority evaluation model defines a flight time urgency function. ; In the formula, For flights At any moment Time urgency rating It is an exponential function. For the current moment, For the planned takeoff time, The attenuation coefficient; Combining the guarantee complexity function Through weighted fusion Calculate priority, where, As the basic weight for flights, For flights At any moment The comprehensive priority score is a weighted fusion of time urgency, guarantee complexity, and basic weights, used to determine whether a flight needs to enter the critical coordination pool for key monitoring. The coefficients are updated regularly based on historical data to achieve adaptive optimization of priority evaluation; When flight time urgency is detected Exceeding the critical threshold The three-tiered judgment mechanism is activated: first, it checks whether the difference between the remaining support time and the planned takeoff time has entered the critical range [15 minutes, 45 minutes]; second, it assesses whether the support status meets the entry conditions; and finally, it confirms the results in conjunction with air traffic control constraints. After all passes, an entry command is generated. The system sends tiered warnings to designated support units and simultaneously activates the resource scheduling engine to calculate the optimal alternative and distribute it to the execution units. If a flight that has undergone manual intervention meets the rollout criteria, its priority is reassessed and added to the main sorting queue for recalculation. The real-time progress monitoring includes: The first step involves collecting multi-source data and constructing a distributed IoT sensor network. Trailers and refueling trucks are equipped with 5G-based positioning modules and status monitoring sensors, and data is transmitted via the MQTT protocol. Simultaneously, a video analytics system using an improved YOLOv8 algorithm identifies key operations such as boarding bridge docking and baggage handling. The advanced YOLOv8 algorithm uses airport cameras to identify key support processes in real time, including boarding bridge docking status, baggage cart loading / unloading actions, and refueling truck operation status. The visual recognition results are converted into support progress data, automatically updating flight support status, replacing manual reporting, and improving data accuracy and real-time performance. The second step is to construct the spatiotemporal state vector, using the flight number as a unique identifier, to build a multidimensional spatiotemporal state vector: ,in, To ensure progress, the report covers the completion status of 17 standard assurance tasks; The resource status component records the location and working status of the guaranteed resources; The table is divided into time components, including key time points planned / expected / actual. To constrain components, record flow control commands and meteorological restriction information; This is used as an event component to record historical anomalous events; The state vector is updated every 1 second, and the completion rate of the guaranteed node is calculated. With delay risk coefficient ; The third step is to design a dynamic threshold triggering mechanism and construct a three-layer cascaded triggering mechanism, including a basic threshold layer, a scene adaptation layer, and a history learning layer, to realize the dynamic generation of thresholds. The fourth step involves generating a reminder instruction. When the remaining coverage time for a flight is below the trigger threshold, the system identifies delayed coverage items and responsible entities, determines priorities based on delay risk, and generates a structured reminder instruction using a template engine. It sends data to the responsible unit through multiple channels, while recording complete operation logs to form a closed-loop tracking system.

[0013] In step S6, the optimization of trailer resource allocation includes: Based on the dynamic flight launch ranking results, an intelligent scheduling engine for trailer resources is constructed. By using IoT positioning technology, the location, operation status and remaining operation time of trailers are monitored in real time. Combined with the completion rate of flight support nodes and taxiing path planning requirements, a multi-objective optimization algorithm is used to construct a trailer-flight matching matrix. Multi-objective optimization algorithms include: constructing a weighted objective function. The expression is: ; In the formula, All of these are dynamically adjusted weighting coefficients. This represents the total task completion time. Total distance traveled. For energy consumption, This refers to the safety margin factor. An improved non-dominated sorting genetic algorithm, NSGA-II, is applied to generate a Pareto optimal solution set. The diversity and convergence of solutions are ensured through fast non-dominated sorting, crowding distance calculation, and elite retention strategies. Finally, a fuzzy comprehensive evaluation mechanism is adopted to select the best matching scheme from the Pareto front, achieving the optimal allocation of trailer resources and flight demand. A dynamic priority strategy is introduced to automatically allocate nearby trailer resources for critically delayed flights and VIP flights, simultaneously rehearse taxiing path conflict risks, and generate trailer scheduling routes without cross-interference. A dynamic priority strategy is introduced to monitor flight status in real time. When a flight is detected to be approaching a critical delay or is marked as a VIP flight, its trailer allocation priority is automatically increased, and the nearest available trailer is forcibly assigned to it. The trailer scheduling of other flights is also re-planned to ensure that important flights are given priority.

[0014] In step S6, the analysis of the causes of delay includes: Based on flight scheduling results and real-time support data stream, a dynamic tracing model for the root causes of delays is constructed. The construction process includes: first, establishing a time point deviation detection network to decompose flight operation trajectories and plans into event sequences and calculate deviation vectors; second, constructing a Bayesian network G(V,E) to identify causal dependencies; and finally, applying spatiotemporal sequence dimensionality reduction technology to form a panoramic view of delay evolution. A multi-level contribution metric framework is employed to identify key delay contributing factors and quantify their contribution weights. This framework utilizes an improved random forest algorithm to screen feature importance I(f). The Shapley method is introduced to decompose delay instance feature attribution. A hierarchical weight model is then constructed. ; In the formula, These are all weighting coefficients for each dimension. Reasons for delay The degree of time-related influence The posterior probability of cause C of delay. The importance of the characteristics of cause C for delay; Calculate the posterior probability and contribution weight of the root causes of delay using Bayesian inference. ; The association rule mining process employs an enhanced Apriori algorithm to extract high-frequency delay patterns; applies graph convolutional networks to model delay propagation; identifies key nodes based on a spatiotemporal attention mechanism; generates interpretable rule sets through decision tree paths, converting them into a visual source tracing report containing a heatmap of delay source distribution, a propagation link graph, a contribution ratio graph, and a trend prediction graph; and automatically pushes targeted optimization suggestions to responsible units.

[0015] In step S7, a dynamic re-entry mechanism is established for flights with resource updates or status changes. When a flight triggers a reordering condition due to updates to the support progress, adjustments to trailer scheduling, or the lifting of airspace restrictions, the priority coefficient is automatically updated based on real-time status parameters. The dynamic reentry mechanism comprises a three-layer trigger-evaluation-execution architecture: the trigger layer uses rule-based event listeners to capture real-time status change signals such as support progress updates, trailer scheduling adjustments, or airspace restriction removal; the evaluation layer uses a state difference detection algorithm to calculate the flight state vector. Compared with the state vector at the previous decision time Distance function When the distance exceeds the adaptive threshold When the reentrancy condition is met, the execution layer calls the priority recalculation function and performs queue reconstruction. Priority coefficients are automatically updated based on real-time status parameters, using a comprehensive scoring formula: ; In the formula, For dynamic time-varying weighting coefficients, For flights exist Feature values ​​of each dimension at any given time; The weight distribution is automatically adjusted according to the operational phase, weather conditions, and resource load, and an ε-greedy strategy is introduced to balance exploration and utilization. The dynamic weight allocation algorithm balances the original flight priorities by introducing a history-current state balance function. Adaptability to the new state The expression is: ; In the formula, It is a time decay function. The weight of the new state is gradually reduced as the waiting time increases, while maintaining the global sorting stability index. When a flight re-entry causes index fluctuations exceeding a preset threshold, a smoothing factor dynamic adjustment mechanism is triggered. This mechanism detects the fluctuation amplitude of the sorting stability index and dynamically adjusts the smoothing coefficient based on the degree of fluctuation, limiting the number of flights that change sorting in a single instance. This ensures a smooth transition in queue adjustments and avoids system oscillations caused by frequent re-sorting.

[0016] Another objective of this invention is to provide a sorting system for airport flight coordination based on multi-source data. This system implements the aforementioned sorting method for airport flight coordination based on multi-source data. The system includes: The multi-source data acquisition module is used to acquire data from multiple data sources related to flights. The multi-source data fusion module is used to fuse collected data from multiple sources to generate a five-dimensional dynamic situation matrix, which reflects the overall status of flights in real time. The module for determining the priority of manual sorting is used to determine whether to perform manual sorting priority processing based on the obtained comprehensive status; if yes, it will make optimization target decisions and intelligent sorting; otherwise, it will jump to the alternative sorting scheme display. The optimization target decision-making and intelligent sorting module is used to optimize target decisions and intelligent sorting based on the results of manual sorting and priority processing, by constructing an intelligent decision-making system for aircraft departure sorting. The alternative sorting scheme display module is used to provide multi-dimensional visual early warnings for real-time flights based on the real-time flight sequence optimization results, with flight bars including adjacent gate prompts and manual intervention functions. The early warning results are displayed as alternative sorting schemes in a visual sorting interface. The module for obtaining the ranking results of flight coordination is used to perform critical coordination pool processing, real-time support progress monitoring, trailer resource allocation optimization, and delay cause analysis on the obtained alternative ranking schemes to obtain the ranking results of flight coordination. The dynamic reentry mechanism module is used to establish a dynamic reentry mechanism for flights with resource updates or status changes based on the ranking results derived from flight coordination, and to re-insert them into the global ranking queue. A dynamic weight allocation algorithm is used to balance the original priority of the flight with its adaptation to the new state. Combining all the above technical solutions, the beneficial effects of this invention are as follows: First, this invention improves airport flight scheduling efficiency by utilizing advanced data analysis techniques and optimization algorithms, thereby reducing flight delays, increasing operational efficiency, and enhancing passenger satisfaction. This invention proposes an innovative flight pushback sequencing method. By constructing a real-time support progress monitoring network and an intelligent trailer resource scheduling system, dynamic support node data and trailer operation trajectory data are synchronously integrated into the sequencing decision model. This reduces the pushback time prediction error from the traditional 15-minute level to within 3 minutes, significantly reducing the risk of cascading delays caused by resource contention and improving ground operation coordination efficiency.

[0017] Secondly, this invention overcomes the technical bottlenecks of protocol heterogeneity and spatiotemporal benchmark differences by constructing a multi-source heterogeneous data fusion engine, enabling synchronous access to flight dynamics, weather radar, and airspace flow control commands. It constructs a five-dimensional dynamic state space, including basic flight data, gate positions, support status, meteorological influencing factors, and airspace gaps, allowing real-time perception of implicit disturbances such as meteorological data and gate aggregation exceeding thresholds. Verified at an international airport, this technology improved flight adjustment response time by four times during thunderstorms and reduced trailer route conflicts by 30%, providing a high-precision data foundation for global optimization.

[0018] Third, this invention designs an adaptive state transition equation and dynamically updates the state space based on real-time data, such as runway occupancy rate and sudden weather signals, supporting second-level replanning. By introducing a conflict detection algorithm and an airspace capacity threshold filtering mechanism, it generates new sequences that meet safety intervals in sudden aircraft failure scenarios, avoiding secondary delays caused by manual adjustments.

[0019] This invention integrates the following indicators into the dynamic programming objective function: delay cost, safety cost, airport on-time departure rate, passenger experience parameters, and airline weights. These indicators are used to assign weighting coefficients to high-priority flights, thereby improving passenger satisfaction. Multiple optimal solutions are generated using a non-dominated sorting algorithm, allowing controllers to select the appropriate solution based on actual needs.

[0020] Fourth, this invention, through multi-source data fusion and dynamic sorting optimization, can significantly improve airport flight on-time departure rates and ground operation efficiency, reduce economic losses caused by flight delays, and optimize the allocation of ground support resources such as tow trucks, thereby reducing airport operating costs. The technical solution can be widely applied to various airports both domestically and internationally, especially large hub airports that urgently need intelligent dispatching systems. It has excellent market prospects and can bring multiple benefits to airport management, airlines, and passengers, promoting the transformation of the civil aviation industry towards a smart operation model.

[0021] Fifth, this invention is the first to achieve real-time fusion and collaborative optimization of flight data, meteorological data, passenger data, air traffic control data, and ground support data in airport flight scheduling, breaking through the technical limitations of traditional single-data-source scheduling. It innovatively proposes a five-dimensional dynamic situation matrix and a dynamic re-entry mechanism, solving the real-time scheduling problem in a multi-vendor heterogeneous data environment, and establishing an intelligent matching system between trailer resources and flight demand. This provides a complete technical solution for the global optimization of airport ground operations, filling a systemic technological gap in this field.

[0022] This invention successfully solves the long-standing problems of resource contention and secondary delays in airport flight scheduling. Traditional methods, relying on manual experience and fixed rules, are ill-suited to cope with dynamically changing and complex operating environments, leading to both idle and wasted trailer resources and flight pushback conflicts. This invention, through real-time status perception and dynamic optimization technology, achieves intelligent coordination of ground support resources and precise control of flight pushback timing, overcoming the challenge of global optimization under multiple constraints. It transforms the passive response of traditional scheduling into proactive prediction and intelligent decision-making, providing a breakthrough technical path for the intelligent upgrading of airport operations management.

[0023] This invention overcomes the technical bias of traditional airport scheduling systems that rely solely on fixed rules and historical experience, and breaks through the cognitive limitation that single-objective optimization cannot simultaneously address multiple objectives such as safety, efficiency, and cost. Traditional views hold that real-time global optimization is impossible in the complex and dynamic environment of airports. This invention, through a dynamic re-entry mechanism and adaptive weight adjustment technology, demonstrates the feasibility of real-time optimized scheduling under multi-source data fusion conditions. Simultaneously, it overturns the traditional perception that manual scheduling is irreplaceable. Through the organic combination of intelligent decision-making systems and human intervention, it achieves a complementary advantage between system intelligence and human experience, opening up a new technological direction for the development of airport scheduling technology. Attached Figure Description

[0024] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure; Figure 1 This is a flowchart of the sorting method for airport flight coordination based on multi-source data provided in this embodiment of the invention. Detailed Implementation

[0025] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0026] The innovation of this invention lies in addressing the technical bottlenecks of traditional airport flight scheduling systems, such as isolated data, inefficient resource scheduling, and insufficient dynamic response capabilities. It innovatively constructs a dynamic situational awareness matrix integrating five data sources: flight data, ground support data, meteorological data, passenger data, and air traffic control data. It also establishes a gate taxiing conflict prediction mechanism based on topological space modeling and a multi-objective optimized intelligent scheduling system for trailer resources. Through a dynamic re-entry mechanism and an adaptive weight adjustment algorithm, it achieves real-time optimization and global coordination of flight pushback sequencing, overcoming the limitations of fixed-rule scheduling. The system integrates a dynamic delay root cause tracing model and a visual decision support interface, reducing flight pushback plan execution deviations from the traditional 15-minute range to within 3 minutes, significantly improving airport ground operation efficiency and resource utilization. This provides a complete technical solution for smart airport construction, realizing a shift from a passive response to a proactive prediction scheduling model.

[0027] Example 1, as Figure 1 As shown, the ranking methods derived from airport flight coordination by integrating multi-source data include: S1 collects data from multiple data sources for flights; these multiple data sources include: flight data sources, ground support data sources, meteorological data sources, passenger data sources, and air traffic control data sources. For example, flight data sources include: The system integrates ACDM system, airline data, etc., to collect flight plans and dynamic information in real time, including but not limited to core parameters such as flight schedule, actual flight time, aircraft type, fuel data, and airline priority weight, and establishes a spatiotemporal state vector with flight number as the unique identifier.

[0028] Ground support data sources include: The system integrates ground service systems, baggage sorting systems, cargo systems, etc., to collect flight support data in real time, including but not limited to flight support status and status data of support resources such as trailers, and achieves second-level data synchronization through the MQTT protocol.

[0029] Meteorological data sources include: real-time access to airport meteorological station observation data, Doppler radar and numerical forecast data through standardized interfaces, including but not limited to parameters such as wind direction and speed, visibility, and thunderstorm impact radius. Real-time data processing latency is ≤200ms, ensuring seamless integration of meteorological data and air traffic control decisions.

[0030] Passenger data sources include: real-time monitoring of passenger security check and boarding status, waiting time, boarding gate density, and other behavioral data through facial recognition gates, passenger departure systems, and video analytics systems; and the use of federated learning to predict group behavior while protecting privacy, thereby optimizing the allocation of passenger service resources.

[0031] Air traffic control data sources include: accessing the air traffic control system to obtain core parameters such as CTOT (Calculated Takeoff Time), COBT (Target Wheelie Time), flow control instructions, runway allocation schemes, and airspace capacity thresholds; dynamically calculating flight priority weights through algorithms; establishing a dynamic airspace capacity model with a 15-minute granularity; and analyzing the route congestion index in real time. When the index exceeds 0.8, the sorting algorithm is recalculated to ensure efficient coordination between air traffic control instructions and ground support.

[0032] S2 integrates the collected data from multiple sources to generate a five-dimensional dynamic situation matrix, which reflects the overall status of flights in real time. A batch processing framework is constructed, and a quaternion interpolation algorithm is used to align the spatiotemporal reference of multi-source data to generate a five-dimensional dynamic situation matrix. The expression is: ; In the formula, For flight data, To provide ground support for data, For meteorological data, For passenger data, This is air traffic control data.

[0033] This matrix is ​​used to reflect the overall status of flights in real time, providing data support for decision-making.

[0034] S3. Based on the obtained comprehensive status, determine whether to perform manual priority sorting; if yes, proceed to step S4; otherwise, proceed to step S5. The system supports commanders in manually adjusting flight sequences through a drag-and-drop interface and manual input of sorting priorities. It implements a technologically improved two-way cognitive collaboration mechanism: the system proactively identifies manual adjustment intentions and recommends the optimal adjustment range, rather than the traditional passive instruction receiving mode. By analyzing user drag-and-drop operation trajectories, click locations, and adjustment magnitudes, the system predicts adjustment intentions, calculates the optimal adjustment range based on a constraint relationship network, and highlights the recommended adjustment interval on the interface. When a manual intervention signal is detected, a priority overriding mechanism is immediately triggered: an enhanced multi-factor authentication framework verifies operation permissions, and the adjusted sequence is written to a version-controlled distributed cache queue. The system verifies multiple pieces of information, including the user's ID card, employee ID, and biometrics. After successful verification, the adjusted sorting data is stored in the distributed cache with timestamps and version numbers, supporting operation rollback and historical record queries.

[0035] The system employs an innovative constraint propagation method to dynamically correct the constraints of subsequent flights: It constructs a flight sequence constraint relationship network, analyzing relationships between flights such as time constraints, airspace constraints, gate constraints, and trailer resource constraints, establishing a directed graph network where nodes represent flights, edges represent constraint relationships, and weights represent constraint strength. The system calculates the impact of manual adjustments on related flights, using the constraint relationship network to calculate the correlation between adjusted flights and other flights. Combining factors such as time window overlap, resource sharing, and priority differences, the system quantifies the impact and assigns impact weights. The principle of minimum adjustment is applied to ensure that only necessary flights are subject to constraint correction. A recursive update technique optimizes the ranking of all affected flights, starting with manually adjusted flights and recursively checking the ranking rationality of their constraint-related flights. If conflicts are found, further adjustments are made until the ranking of all related flights meets the constraint conditions. The system calculates an adjustment necessity score for each flight, adjusting only flights with scores exceeding a threshold, prioritizing maintaining the ranking of most flights unchanged, and minimizing the overall adjustment range. Simultaneously, it maintains the stability and rationality of the global ranking. This invention is the first to propose the "minimum impact ripple" principle, significantly reducing chain constraint conflicts compared to traditional manual intervention methods. The principle of "minimizing impact" includes four aspects: localized adjustment (adjusting only necessary flights), gradual spread (adjusting relevant flights in stages), impact boundary control (limiting the scope of adjustment spread), and stability protection (keeping the order of most flights unchanged).

[0036] S4, based on the results of manual sorting and priority processing, optimizes target decision-making and intelligent sorting by constructing an intelligent decision-making system for aircraft departure sorting; This invention constructs an intelligent decision-making system for aircraft departure sequencing through multi-dimensional parameter coupling and dynamic weight adjustment, and establishes a comprehensive scoring function as follows: ; In the formula, As a weighting factor for delay costs, For safety cost weighting coefficient, This is a weighting coefficient for the airport's on-time departure rate. This is a weighting coefficient for the airport's on-time departure rate. For passenger experience parameters, weighting coefficients For airlines, the weighting coefficient is... To address delay costs, a delay cost matrix is ​​constructed, categorized by aircraft type and airline. Statutory compensation standards and carbon emission conversion cost indicators are introduced, and the cost coefficient is dynamically adjusted by training the model using airline operational data. For safety and cost reasons, the standard computer type combination interval is affected by time slot limitations; The airport on-time departure rate parameter is defined as a parameter affecting the airport's overall departure rate. For passenger experience parameters, passenger impact parameters; For airlines, consider factors such as weight, market share, historical on-time rate, historical departure status, and VIP flight marking; for example, if the top three airlines with the most monthly flight departure delays are delayed, coordinate with the airlines to adjust for more than 60 minutes under pre-tactical conditions, and ensure that flights with more than 60 minutes of flight time adjustment are given higher priority in the queue.

[0037] Substituting the real-time status data of each flight into the above formula, a comprehensive score for each flight is calculated. The priority ranking of flights is determined based on the scores, forming a baseline intelligent launch ranking scheme. Based on the baseline ranking scheme, real-time adjustments are made when the following situations occur: airlines temporarily requesting priority adjustments, VIP flights urgently jumping the queue, sudden equipment failures affecting specific flights, temporary changes in air traffic control instructions, and sudden changes in weather conditions. The system recalculates the priority of affected flights based on temporary changes and dynamically adjusts the ranking sequence to ensure that the ranking scheme always adapts to the actual operational needs on site.

[0038] S5, based on the real-time flight sequence optimization results, performs multi-dimensional visual early warnings on real-time flights with flight bars that include adjacent gate prompts and manual intervention functions, and displays the early warning results in a visual sorting interface to present alternative sorting schemes. For example, flights are displayed in the form of flight bars. In addition to basic flight data, the flight bars also have adjacent gate prompts and manual intervention functions, and reserve dynamic adjustment functions for special circumstances. The manual intervention function allows controllers to directly modify the position of a flight in the sorting queue by dragging or typing in special circumstances (such as VIP flights, emergency medical flights, sudden malfunctions, etc.). The system will automatically recalculate the sorting of affected flights and update the conflict risk assessment in real time. This function has the highest priority and can override the system's intelligent sorting results, ensuring a rapid response to on-site needs in emergency situations. At the same time, the system will record all manual adjustment operations for traceability and analysis.

[0039] Its core function is to preserve the controller's final decision-making authority and emergency response capabilities based on automated and intelligent sequencing.

[0040] Among them, the adjacent aircraft stand prompt function adopts a topological space modeling and dynamic risk assessment system to predict and warn of potential taxiing conflicts between aircraft stands.

[0041] For example, a topology space modeling and dynamic risk assessment system can be used to predict and warn of potential taxiing conflicts between aircraft positions, including: S5.1 Identification of adjacent camera positions.

[0042] System construction station topology diagram Among them, vertex set Represents the machine station node and edge set. Represents the intersection relationships of potential gliding paths; each edge Carrying weight vector , These represent physical distance, gliding angle difference, path intersection count, and historical conflict frequency, respectively; the historical conflict frequency is dynamically updated through gliding trajectory data to achieve adaptive optimization of topology relationships; Construct a conflict risk assessment model and define the aircraft position pair risk factor : ; In the formula, All are dynamically weighted coefficients. This is a spatial correlation function. This is a time window overlap function. For aircraft characteristic functions, For meteorological influence functions, For the camera position The weight vector, For camera position and camera position At any moment The overlap of the launch time windows For camera position At any moment Aircraft characteristic parameters, For camera position At any moment Aircraft characteristic parameters, For a moment Meteorological state parameters; The system uses a sliding time window to predict the risk of gliding operations within the next T minutes, forming a spatiotemporal risk heat matrix. When the risk coefficient exceeds the threshold, an early warning mechanism is triggered.

[0043] S5.2 Visualized Early Warning.

[0044] The system is based on the calculated risk coefficient To achieve multi-dimensional visual early warning: Color coding: The risk coefficient is mapped to the HSV color space. Low risk is displayed as green, medium risk as yellow, high risk as orange, and critical risk as red with a flashing effect to improve visual salience.

[0045] Interactive Topology Map: Dynamic topology relationship layers are overlaid on flight bars. The strength of association and the risk of conflict are represented by the thickness, color and solidity of the lines. It supports multi-scale interactive operations and can adapt to decision-making scenarios of varying complexity.

[0046] 3D Scene Reconstruction: For complex aircraft stand layouts, it provides 3D scene reconstruction function, overlaying aircraft movement trajectories and conflict hotspots to help commanders intuitively understand the causes and impact range of potential conflicts.

[0047] Based on the detected conflict risks, the system automatically generates alternative elimination sorting schemes and displays the risk changes brought about by the scheme adjustment in the form of a difference heatmap. At the same time, it works in conjunction with modules such as trailer resource allocation and critical coordination pool handling in step S6 to minimize the global conflict risk.

[0048] The visual sorting interface displays a countdown timer for the flight bar based on the number of flights and control positions, with color-changing indicators to alert ground support and apron control to pay attention and ensure that the flight is pushed back and taken off within the expected release time or the restricted time.

[0049] Flight volume and control positions: Flight volume determines the density of gate usage, and control positions determine the workload of controllers. These two factors affect the probability of taxiing conflicts between adjacent gates and the control response capability. The system dynamically adjusts the sensitivity of gate conflict risk assessment based on these conditions.

[0050] The flight bar displays a countdown to the estimated departure time and the restricted time. The countdown display is combined with the gate conflict warning: when the risk of an adjacent gate is high and the countdown is approaching, the flight bar color will comprehensively display the time urgency and the conflict risk level, reminding the controller to prioritize the pushback or adjustment of this flight.

[0051] S6: For the obtained alternative launch sorting schemes, perform critical coordination pool processing, real-time support progress monitoring, trailer resource allocation optimization, and delay cause analysis to obtain the flight coordination launch sorting results; The critical coordination pool processing includes: employing a dynamic triggering mechanism based on multimodal data fusion; and constructing a priority evaluation model based on a time window decay function by real-time collection of flight support progress, air traffic control instructions, and resource availability status. This model defines a flight time urgency function. ; In the formula, For flights At any moment Time urgency rating It is an exponential function. For the current moment, For the planned takeoff time, The attenuation coefficient; Combining the guarantee complexity function Through weighted fusion Calculate priority, where, As the basic weight for flights, For flights At any moment The comprehensive priority score is a weighted fusion of time urgency, guarantee complexity, and basic weights, used to determine whether a flight needs to enter the critical coordination pool for key monitoring; the system updates each coefficient regularly based on historical data to achieve adaptive optimization of priority assessment.

[0052] When flight time urgency is detected Exceeding the critical threshold (Default 0.85), activate the three-layer judgment mechanism: First, check whether the difference between the remaining support time and the planned takeoff time enters the critical range [15 minutes, 45 minutes]; second, assess whether the support status meets the entry conditions; finally, confirm in conjunction with air traffic control constraints. After all passes, an entry command is generated. The system sends tiered warnings (reminder, attention, emergency) to designated support units, and simultaneously activates the resource scheduling engine to calculate the optimal alternative and distribute it to the execution units. If a flight that has undergone manual intervention meets the pushback conditions, its priority is reassessed and added to the main sorting queue for recalculation. For example, real-time progress monitoring includes: By integrating data from the ground service system, real-time monitoring and early warning management of the entire airport ground support process were achieved. Its execution process encompasses the following four core steps: The first step is the collection of multi-source data: This system constructs a distributed Internet of Things sensor network, in which mobile devices such as trailers and refueling trucks are equipped with positioning modules and status monitoring sensors based on 5G technology, and data is transmitted through the MQTT protocol; at the same time, a video analysis system using an improved YOLOv8 algorithm identifies key operations such as boarding bridge docking and baggage loading and unloading.

[0053] The second step is the construction of the spatiotemporal state vector: the system uses the flight number as a unique identifier to construct a multidimensional spatiotemporal state vector. ,in, To ensure progress, the report covers the completion status of 17 standard assurance tasks; The resource status component records the location and working status of the guaranteed resources; The table is divided into time components, including key time points planned / expected / actual. To constrain components, record flow control commands and meteorological restriction information; This is used as an event component to record historical anomalous events; The system updates the state vector every 1 second and calculates the completion rate of the guaranteed nodes. With delay risk coefficient Ensure node completion rate. : Count the number of tasks that have been completed out of the 17 standard guarantee tasks, and calculate the completion rate. If 12 tasks have been completed, then CR = 12 / 17 = 70.6%.

[0054] Delay risk factor The risk score is calculated by taking into account factors such as the remaining time for ensuring the mission, the number of unfinished critical tasks, and the availability of resources. The higher the score, the greater the risk of delay.

[0055] The third step involves the design of a dynamic threshold triggering mechanism: The system constructs a three-layer cascaded triggering mechanism, including a basic threshold layer, a scenario adaptation layer, and a historical learning layer, to achieve dynamic generation of thresholds. The basic threshold layer sets a fixed baseline triggering threshold based on flight type; the scenario adaptation layer adjusts the threshold in real time based on current weather conditions, flight density, equipment status, etc.; the historical learning layer optimizes the threshold setting based on historical support data and delay patterns using machine learning algorithms. The three layers collectively calculate the final dynamic triggering threshold.

[0056] The fourth step is the generation of the reminder instruction: When the remaining coverage time for a flight is lower than the trigger threshold, the system will identify the delayed coverage items and responsible units, determine the priority based on the delay risk, and generate a structured reminder instruction M(f,t,Pi,Di) through a template engine. The system compares the planned time and actual progress of 17 coverage tasks, identifies overdue coverage items, queries the responsible unit information, determines the reminder priority level based on the delay risk coefficient, and automatically fills in information such as flight number, delayed items, responsible unit, and priority through a preset template to generate a standardized reminder instruction and send it through multiple channels. It sends the instruction to the responsible unit through multiple channels (App push, SMS, voice) and records a complete operation log to form a closed-loop tracking system.

[0057] For example, trailer resource allocation optimization includes: Based on the dynamic flight launch ranking results, an intelligent scheduling engine for trailer resources is constructed. By utilizing IoT positioning technology to monitor the location, operational status, and remaining operational time of trailers in real time, and combining the completion rate of flight support nodes with taxiing path planning requirements, this invention employs a multi-objective optimization algorithm to construct a trailer-flight matching matrix.

[0058] The implementation strategies of this algorithm include: First, construct the weighted objective function. In the formula, All of these are dynamically adjusted weighting coefficients. This represents the total task completion time. Total distance traveled. For energy consumption, This refers to the safety margin factor. Secondly, an improved non-dominated sorting genetic algorithm (NSGA-II) is applied to generate a Pareto optimal solution set. Fast non-dominated sorting, crowding distance calculation, and elite retention strategies ensure the diversity and convergence of solutions. Finally, a fuzzy comprehensive evaluation mechanism is employed to select the optimal matching scheme from the Pareto front, achieving the optimal allocation of trailer resources to flight demand. The system introduces a dynamic priority strategy to automatically allocate nearby trailer resources to critically delayed flights and VIP flights, simultaneously rehearsing taxiing path conflict risks, and generating trailer scheduling routes without cross-interference, effectively eliminating ineffective waiting caused by resource contention.

[0059] For example, delay cause analysis includes: Based on flight scheduling results and real-time support data streams, a dynamic tracing model for the root causes of delays is constructed. The construction process includes: first, establishing a time-point deviation detection network to decompose flight trajectories and plans into event sequences and calculate deviation vectors; second, constructing a Bayesian network G(V,E) to identify causal dependencies; and finally, applying spatiotemporal sequence dimensionality reduction techniques to form a panoramic view of delay evolution.

[0060] The system employs a multi-level contribution quantification framework to identify key delay causes and quantify their contribution weights. This framework uses an improved random forest algorithm to screen feature importance I(f); introduces the Shapley method to decompose delay instance feature attribution; and constructs a hierarchical weight model. ： ; In the formula, These are all weighting coefficients for each dimension. Reasons for delay The degree of time-related influence The posterior probability of cause C of delay. The importance of the characteristics of cause C for delay; Calculate the posterior probability and contribution weight of the root causes of delay using Bayesian inference. ; The association rule mining process employs an enhanced Apriori algorithm to extract high-frequency delay patterns; applies graph convolutional networks to model delay propagation; identifies key nodes based on a spatiotemporal attention mechanism; and generates interpretable rule sets through decision tree paths, converting them into a visual source tracing report containing a heatmap of delay source distribution, a propagation link graph, a contribution ratio graph, and a trend prediction graph. The system automatically pushes targeted optimization suggestions to responsible units, supporting continuous improvement of operational quality.

[0061] S7: Based on the sorting results obtained from flight coordination, a dynamic re-entry mechanism is established for flights with resource updates or status changes, and the flights are re-inserted into the global sorting queue; through a dynamic weight allocation algorithm, the original priority of the flights is balanced with the adaptability of the new status.

[0062] When a flight triggers a reordering condition due to updates to the support progress, adjustments to trailer scheduling, or the lifting of airspace restrictions, the priority coefficient is automatically updated based on real-time status parameters, and the flight with updated resources or changed status is re-inserted into the global sorting queue. The dynamic re-entry mechanism proposed in this invention comprises a three-layer trigger-evaluation-execution architecture: the trigger layer captures state change signals such as progress updates, trailer scheduling adjustments, or airspace restriction removal in real time through a rule-based event listener; the evaluation layer uses a state difference detection algorithm to calculate the distance function d(S(t),S(t-Δt)) between the flight state vector S(t) and the state vector S(t-Δt) at the previous decision time, and confirms that the re-entry condition is met when the distance exceeds the adaptive threshold θ(t); the execution layer calls the priority recalculation function and performs queue reconstruction operation.

[0063] The system automatically updates priority coefficients based on real-time status parameters, using the comprehensive scoring formula P(f,t)=∑λ i (t)·f i (t), where f i (t) represents the feature values ​​of flight f at time t in various dimensions, λ i (t) represents the dynamic time-varying weighting coefficient. The system automatically adjusts the weight distribution based on the operational phase, weather conditions, and resource load, and introduces an ε-greedy strategy to balance exploration and utilization, ensuring that the priority coefficient accurately reflects the current comprehensive support status of the flight.

[0064] The dynamic weight allocation algorithm introduces a history-current state balance function B(f,t)=α·P o (f)+(1-α)·P n (f,t) is used to balance the original flight priority P. o (f) Fit to the new state P n (f,t), where α is the time decay function α(t)=exp(-γ·t), which gradually reduces the weight of the new state as the waiting time increases to prevent queue oscillations caused by frequent state updates of low-priority flights. The system also maintains a global sorting stability index S(Q). When the index fluctuation caused by a flight re-entry exceeds a preset threshold, a smoothing factor dynamic adjustment mechanism is triggered to ensure the overall smooth evolution of the queue structure.

[0065] As demonstrated by the above embodiments, this invention significantly improves airport flight on-time departure rates and ground support efficiency through multi-source data fusion and intelligent optimization decision-making, effectively reducing the risk of secondary delays caused by sudden disturbances, and achieving multi-objective synergistic optimization of safety intervals, fuel economy, and passenger experience. The system rapidly generates compliant dispatching plans under complex weather conditions and equipment failure scenarios, simultaneously optimizing trailer resource utilization efficiency and taxiway path planning, reducing carbon emissions and manual coordination workload. Furthermore, it enhances the transparency of air traffic control decisions through a 3D visualization interface, providing passengers with smoother travel services and comprehensively promoting the transformation of airport operations towards intelligence and green practices.

[0066] As can be seen from the above embodiments, the present invention has the following advantages: Flight Pushback Sequencing: This invention addresses the inaccuracy issues caused by the dynamic nature of ground support and inefficient resource scheduling in traditional departure sequencing mechanisms. It proposes a flight pushback sequencing method based on dynamic data fusion. By constructing a real-time support progress monitoring network, integrating flight support node status and trailer resource trajectory data, and combining meteorological influence factors and airspace capacity attenuation coefficients, a dynamic flight pushback priority sequence is generated. The system employs an intelligent matching algorithm to optimize trailer scheduling paths, eliminating invalid waiting caused by resource contention. Simultaneously, a taxiing conflict pre-detection mechanism is introduced to ensure that the pushback sequence simultaneously meets safety interval and operational efficiency requirements. This technology reduces the pushback time prediction error from the traditional 15-minute level to within 3 minutes. Furthermore, a dynamic command push system enables collaborative operation between support units and air traffic control, ultimately forming a closed-loop control system of "status awareness - intelligent decision-making - precise execution," significantly improving the ground operation efficiency and resource utilization of departing flights.

[0067] Data Fusion Engine: This invention utilizes standardized protocol interfaces to access flight operation data, meteorological data, airspace status data, ground support data, and passenger behavior data in real time. It also performs outlier correction and key dynamic feature extraction on the real-time data stream. Through algorithms, multi-dimensional data is mapped to a unified coordinate system, generating a global situational matrix with time granularity. This matrix integrates five dynamic parameters: basic flight attributes, gate topology relationships, quantitative indicators of support progress, meteorological impact levels, and airspace clearance capacity. This achieves high-precision fusion and millisecond-level updates of multi-source data, providing a highly timely and reliable data foundation for global optimization decisions.

[0068] Adaptive Dynamic Programming Model: This invention addresses the response lag and secondary delays caused by sudden disturbances during aircraft ground operations by proposing a dynamic adaptive real-time response technology. By constructing an adaptive decision-making model based on reinforcement learning, it collects dynamic parameters such as meteorological change warning signals and airspace capacity decay index in real time, driving millisecond-level updates and replanning of the state space. The system integrates a high-precision conflict detection algorithm and a dynamic airspace capacity threshold filtering mechanism. In emergency scenarios such as sudden aircraft malfunctions, it automatically generates alternate landing sequences and trailer scheduling schemes that meet safety separation standards, simultaneously optimizing taxiing routes and resource allocation strategies. This technology breaks through the traditional manual intervention mode, achieving ten-second-level disturbance response and autonomous decision-making, effectively eliminating the risk of secondary delays, and significantly improving operational resilience and handling efficiency under complex dynamic constraints.

[0069] Multi-Objective Collaborative Optimization Engine: This invention addresses the imbalance between safety, efficiency, and service quality caused by traditional single-objective optimization models by proposing a multi-dimensional collaborative optimization method. By constructing a dynamic programming objective function, it simultaneously integrates safety interval standards, fuel economy indicators, and passenger experience parameters, and assigns dynamic weighting coefficients to high-priority flights to achieve differentiated support strategies. The system employs advanced multi-objective optimization algorithms to generate Pareto optimal solutions, combined with a 3D visualization decision-making interface, allowing controllers to autonomously weigh safety margins, clearance efficiency, and service quality based on real-time operational conditions. This technology overcomes the limitations of single-indicator optimization, effectively reducing fuel consumption and improving passenger satisfaction while ensuring operational safety, thus driving the transformation of airport ground operations towards a multi-objective collaborative optimization model.

[0070] For example, Figure 1 This invention relates to the principle of a sorting method derived from airport flight coordination that integrates multi-source data, as provided in this embodiment.

[0071] Example 2: This invention provides a sorting system for airport flight coordination that integrates multi-source data. The system includes: The multi-source data acquisition module is used to acquire data from multiple data sources related to flights. The multi-source data fusion module is used to fuse collected data from multiple sources to generate a five-dimensional dynamic situation matrix, which reflects the overall status of flights in real time. The module for determining the priority of manual sorting is used to determine whether to perform manual sorting priority processing based on the obtained comprehensive status; if yes, it will make optimization target decisions and intelligent sorting; otherwise, it will jump to the alternative sorting scheme display. The optimization target decision-making and intelligent sorting module is used to optimize target decisions and intelligent sorting based on the results of manual sorting and priority processing, by constructing an intelligent decision-making system for aircraft departure sorting. The alternative sorting scheme display module is used to provide multi-dimensional visual early warnings for real-time flights based on the real-time flight sequence optimization results, with flight bars including adjacent gate prompts and manual intervention functions. The early warning results are displayed as alternative sorting schemes in a visual sorting interface. The module for obtaining the ranking results of flight coordination is used to perform critical coordination pool processing, real-time support progress monitoring, trailer resource allocation optimization, and delay cause analysis on the obtained alternative ranking schemes to obtain the ranking results of flight coordination. The dynamic re-entry mechanism module is used to establish a dynamic re-entry mechanism for flights with resource updates or status changes based on the ranking results obtained from flight coordination, and to re-insert them into the global ranking queue; through a dynamic weight allocation algorithm, the original priority of the flight is balanced with the adaptability of the new state.

[0072] To verify the effectiveness of the proposed "Sorting Method and System for Airport Flight Coordination," a six-month system application test was conducted at an international airport. Comprehensive simulation test schemes were designed for different operational scenarios. By comparing and analyzing the differences between traditional methods and the method of this invention in various key performance indicators, the actual application effect of the system was objectively evaluated.

[0073] 1. Experimental environment and data acquisition.

[0074] The simulation test of this invention was conducted at an international airport for six months (2024), covering different operating hours including weekdays, weekends, and holidays, involving more than 80,000 flights. During the test, various weather conditions were encountered, including clear skies, rain, fog, and thunderstorms, providing comprehensive verification of the system's performance under different weather environments. The test platform employed a high-performance computing server cluster, configured with multi-core CPU processors, large-capacity memory and storage, connected via the airport's high-speed internal network, with an average latency controlled within 5ms. Multiple edge computing terminals were deployed at various support positions throughout the airport, constructing a complete experimental technical environment.

[0075] In terms of data acquisition, the system accesses flight plans and dynamic information in real time from the ACDM system and airline operation databases, integrates status information of support nodes provided by multiple ground service systems, connects with meteorological data from airport weather stations, Doppler radar, and numerical weather prediction systems, collects passenger dynamics in real time through security check systems and boarding gate video systems, connects to the ATFM system and flow control management system to obtain air traffic control data updated minute by minute, and tracks the location and status of trailers in real time through a trailer GPS positioning system. This multi-source data fusion acquisition mode provides the system with comprehensive, timely, and accurate operational status information, providing a solid data foundation for subsequent optimization decisions.

[0076] 2. Verification test of the effect in a normal operating scenario.

[0077] 2.1 Test protocol.

[0078] To verify the effectiveness of the system under normal operating conditions, we selected operational data from a period of continuous clear weather for analysis. We compared the differences between the traditional fixed-rule sorting method and the dynamic data fusion sorting method of this invention in key indicators such as flight on-time departure rate, ground dwell time, and trailer utilization efficiency.

[0079] 2.2 The test results are shown in Table 1 and Table 2.

[0080] Table 1 Comparison of Effects in Typical Operating Scenarios

[0081] Table 2 Comparison of Flight Departure Sequencing Execution Efficiency

[0082] Analysis shows that, under normal operating conditions, the system of this invention improved the airport's on-time departure rate from 89.23% to 89.87%; the average ground dwell time decreased from 18.7 minutes to 13.4 minutes, a reduction of 28.3%; trailer utilization efficiency increased by 42.4%; and the resource conflict rate was reduced by 75.2%. Particularly during peak hours, the deviation in flight pushback plan execution was controlled within 3.6 minutes, a reduction of 76.5% compared to traditional methods. This verifies that the present invention, by constructing a real-time support progress monitoring network and an intelligent trailer resource scheduling system, effectively improves ground operation coordination efficiency.

[0083] 3. Verification test of effects in special weather scenarios. 3.1 Test protocol.

[0084] To verify the effectiveness of this invention in dealing with special weather conditions, we selected operational data under special meteorological conditions such as thunderstorms, heavy fog, and strong winds for analysis, and compared the performance of traditional single data source methods and the multi-source fusion method of this invention on various key indicators.

[0085] 3.2 The test results are shown in Table 3.

[0086] Table 3 Comparison of Effects in Special Weather Scenarios

[0087] Analysis results show that the system of this invention improved the on-time departure rate under various special weather conditions, especially under thunderstorm conditions, where the on-time departure rate increased from 68.4% to 72.3%, an improvement of 5.7%. Simultaneously, the average ground dwell time was significantly reduced, and trailer utilization efficiency was significantly improved. Most importantly, the system shortened the weather recovery time to normal operation from an average of 4.6 hours to 1.8 hours, a reduction of 60.9%, verifying the effectiveness of this invention in breaking down data silos and constructing a global optimization decision-making foundation through multi-source data fusion.

[0088] 4. Emergency scenario effect verification test. 4.1 Test protocol.

[0089] To verify the ability of this invention to respond to emergencies, we selected three typical emergency scenarios for analysis: temporary airspace restrictions, sudden flow control, and temporary closure of ground facilities. We compared the performance differences between traditional methods and the method of this invention in terms of emergency response speed and secondary delay control.

[0090] 4.2 The test results are shown in Table 4.

[0091] Table 4 Comparison of Emergency Scenarios

[0092] Analysis shows that the system of this invention performs well in emergency scenarios. Under temporary airspace restrictions, the on-time departure rate increased from 74.5% to 78.2%, and the proportion of secondary delays decreased by 67.6%. Under sudden flow control measures, the on-time departure rate increased from 70.6% to 74.8%, and the resource conflict rate decreased by 75.5%. Under temporary closure of ground facilities, the resource reconfiguration time was reduced from 12.3 minutes to 3.6 minutes, a decrease of 70.7%. These results verify the system's ability to achieve rapid response and accurate decision-making under complex dynamic constraints.

[0093] 5. Analysis of the overall application effect of the system. During its six-month operation at an international airport, the system of this invention demonstrated comprehensive application effects, as shown in Table 5.

[0094] Table 5. Overall System Application Effect

[0095] Especially during a special scenario test involving a severe thunderstorm during the trial period, the advantages of this system further include: A large number of flights were dynamically reordered in a short period of time.

[0096] Trailer utilization efficiency increased to 92.4%, which is 5.6% higher than normal operation.

[0097] The resource conflict rate was reduced to 4.2%, a decrease of 4.1% compared to normal operation.

[0098] The time to restore normal operation was 62.5% shorter than in similar historical situations.

[0099] 6. Through comprehensive simulation experiments and practical application verification of the "Sorting Method and System for Airport Flight Coordination" of this invention, the following results were obtained: In normal operating scenarios, the system of this invention improves the on-time departure rate of airport flights, reduces the average ground dwell time, enhances trailer utilization efficiency, reduces the conflict rate of support resources, and verifies the effectiveness of the innovative flight pushback sequencing method.

[0100] In special weather scenarios, the system of this invention breaks down data silos by fusing multi-source data, effectively copes with adverse weather conditions such as thunderstorms, heavy fog, and strong winds, improves the on-time departure rate of flights, and shortens the time to restore normal operation.

[0101] In emergency scenarios, the system of this invention demonstrates rapid response and accurate decision-making capabilities. It effectively reduces the risk of secondary delays, improves resource utilization efficiency, and enhances the operational resilience of the system in the face of temporary airspace restrictions, sudden flow control, and temporary closure of ground facilities.

[0102] The system has achieved significant overall results, with an increase of 0.72% in the airport's on-time departure rate, a reduction of 32.3% in average ground dwell time, an increase of 34.2% in trailer utilization efficiency, a reduction of 71.1% in resource conflict rate, and monthly savings of approximately 284 tons of fuel and a reduction of approximately 892 tons of carbon emissions.

[0103] The successful implementation of this invention at an international airport not only verified the effectiveness and advancement of its core technology, but also provided an innovative solution for the construction of smart airports in civil aviation.

[0104] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A sorting method derived from airport flight coordination by integrating multi-source data, characterized in that, The method includes the following steps: S1 collects data from multiple data sources for flights; these multiple data sources include: flight data sources, ground support data sources, meteorological data sources, passenger data sources, and air traffic control data sources. S2 integrates the collected data from multiple sources to generate a five-dimensional dynamic situation matrix, which reflects the overall status of flights in real time. S3. Based on the obtained comprehensive status, determine whether to perform manual priority sorting; if yes, proceed to step S4; otherwise, proceed to step S5. S4, based on the results of manual sorting and priority processing, optimizes target decision-making and intelligent sorting by constructing an intelligent decision-making system for aircraft departure sorting; S5, based on the real-time flight sequence optimization results, performs multi-dimensional visual early warnings on real-time flights with flight bars that include adjacent gate prompts and manual intervention functions, and displays the early warning results in a visual sorting interface to present alternative sorting schemes. S6: For the obtained alternative launch sorting schemes, perform critical coordination pool processing, real-time support progress monitoring, trailer resource allocation optimization, and delay cause analysis to obtain the flight coordination launch sorting results; S7: Based on the sorting results obtained from flight coordination, a dynamic re-entry mechanism is established for flights with resource updates or status changes, and the flights are re-inserted into the global sorting queue; through a dynamic weight allocation algorithm, the original priority of the flights is balanced with the adaptability of the new status.

2. The sorting method for airport flight coordination based on multi-source data as described in claim 1, characterized in that, In step S2, a five-dimensional dynamic situation matrix is ​​generated, including: A batch processing framework is constructed, and a quaternion interpolation algorithm is used to align the spatiotemporal reference of multi-source data to generate a five-dimensional dynamic situation matrix. The expression is: ； In the formula, For flight data, To provide ground support for data, For meteorological data, For passenger data, This is air traffic control data.

3. The sorting method for airport flight coordination based on multi-source data as described in claim 1, characterized in that, In step S3, based on the obtained overall status, it is determined whether to perform manual priority sorting, including: The improved two-way cognitive collaboration mechanism is achieved by manually adjusting flight sequences through a drag-and-drop interactive interface and manual input of sorting priorities. The improved two-way cognitive collaboration mechanism includes: actively identifying manual adjustment intentions and recommending the optimal adjustment range; triggering a priority overriding mechanism when a manual intervention signal is detected; verifying operation permissions through an enhanced multi-factor authentication framework; and writing the adjusted sequence into a distributed cache queue with version control capabilities. The constraint propagation method is used to dynamically correct the constraints of subsequent flights: a flight sequence constraint relationship network is constructed, the impact of manual adjustments on related flights is calculated, the minimum adjustment principle is applied to only correct the constraints of necessary flights, and the sorting position of all affected flights is optimized through recursive update technology.

4. The sorting method for airport flight coordination based on multi-source data as described in claim 1, characterized in that, In step S4, by constructing an intelligent decision-making system for aircraft departure sequencing, optimization target decisions and intelligent sequencing are performed, including: By coupling multi-dimensional parameters and adjusting dynamic weights, an intelligent decision-making system for aircraft departure sequencing is constructed. A comprehensive scoring function is established, and multiple key factors are considered through weighted summation to determine the priority of flight sequencing. The expression is as follows: ； In the formula, As a weighting factor for delay costs, For safety cost weighting coefficient, This is a weighting coefficient for the airport's on-time departure rate. This is a weighting coefficient for the airport's on-time departure rate. For passenger experience parameters, weighting coefficients For airlines, the weighting coefficient is... To address delay costs, a delay cost matrix is ​​constructed, categorized by aircraft type and airline. Statutory compensation standards and carbon emission conversion cost indicators are introduced, and the cost coefficient is dynamically adjusted by training the model using airline operational data. For safety and cost reasons, the standard computer type combination interval is affected by time slot limitations; The airport on-time departure rate parameter is defined as a parameter affecting the airport's overall departure rate. For passenger experience parameters, passenger impact parameters; For airline weight, market share, historical on-time rate, historical departure status, VIP flight marking; Substituting the real-time status data of each flight into the above formula, a comprehensive score for each flight is calculated. The priority ranking of flights is determined based on the scores, forming a baseline intelligent launch ranking scheme. Based on the baseline ranking scheme, real-time adjustments are made when the following situations occur: airlines temporarily requesting priority adjustments, VIP flights urgently jumping the queue, sudden equipment failures affecting specific flights, temporary changes in air traffic control instructions, and sudden changes in weather conditions. The system recalculates the priority of affected flights based on temporary changes and dynamically adjusts the ranking sequence to ensure that the ranking scheme always adapts to the actual operational needs on site.

5. The sorting method for airport flight coordination based on multi-source data as described in claim 1, characterized in that, In step S5, the adjacent aircraft stand prompting function uses a topology space modeling and dynamic risk assessment system to predict and warn of potential taxiing conflicts between aircraft stands. A topology-based spatial modeling and dynamic risk assessment system is used to predict and provide early warnings of potential taxiing conflicts between aircraft positions, including: S5.1 Identification of Adjacent Camera Positions: System construction station topology diagram Among them, vertex set Represents the machine station node and edge set. Represents the intersection relationships of potential gliding paths; each edge Carrying weight vector , These represent physical distance, gliding angle difference, path intersection count, and historical conflict frequency, respectively; the historical conflict frequency is dynamically updated through gliding trajectory data to achieve adaptive optimization of topology relationships; Construct a conflict risk assessment model and define the aircraft position pair risk factor : ； In the formula, All are dynamically weighted coefficients. This is a spatial correlation function. This is a time window overlap function. For aircraft characteristic functions, For meteorological influence functions, For the camera position The weight vector, For camera position and camera position At any moment The overlap of the launch time windows For camera position At any moment Aircraft characteristic parameters, For camera position At any moment Aircraft characteristic parameters, For a moment Meteorological state parameters; Using a sliding time window, the risk of gliding operations within the next T minutes is predicted, forming a spatiotemporal risk heat matrix. When the risk coefficient exceeds the threshold, an early warning mechanism is triggered. S5.2 Visual Early Warning; Based on the calculated risk coefficient To achieve multi-dimensional visual early warning, including: Color coding maps the risk level to the HSV color space, with low risk displayed as green, medium risk as yellow, high risk as orange, and critical risk as red with an added flashing effect; Interactive topology map, which overlays dynamic topology relationship layers on flight bars, uses line thickness, color, and solidity to represent the strength of association and risk of conflict; 3D scene reconstruction: For complex aircraft stand layouts, it provides 3D scene reconstruction function, overlays aircraft movement trajectory and conflict hotspot areas to obtain potential conflict causes and impact range; Simultaneously, based on the detected conflict risks, alternative pushback sequencing schemes are automatically generated. After detecting taxiing conflict risks between aircraft stands, the sequencing formula is automatically recalculated to generate multiple sequencing combinations that avoid conflicts. Each scheme adjusts the pushback sequence of different flights to ensure that taxiing paths do not intersect or conflict. The risk changes brought about by the scheme adjustments are displayed in the form of a difference heatmap, comparing the risk coefficients of the original scheme and the alternative schemes, and using color depth to indicate the degree of risk change: green indicates reduced risk, red indicates increased risk, and the darker the color, the greater the change, intuitively showing the risk impact of each adjustment scheme. In conjunction with the trailer resource allocation and critical coordination pool handling in step S6, the overall conflict risk is minimized. The visual sorting interface displays a countdown to the estimated departure time and restricted time based on the number of flights and control positions, with color changes indicating that ground support and apron control should push back and take off within the estimated departure time or restricted time.

6. The sorting method for airport flight coordination based on multi-source data as described in claim 1, characterized in that, In step S6, the critical coordination pool processing includes: employing a dynamic triggering mechanism based on multimodal data fusion, and constructing a priority evaluation model based on a time window decay function by real-time collection of flight support progress, air traffic control instructions, and resource availability status. This priority evaluation model defines a flight time urgency function. ； In the formula, For flights At any moment Time urgency rating It is an exponential function. For the current moment, For the planned takeoff time, The attenuation coefficient; Combining the guarantee complexity function Through weighted fusion Calculate priority, where, As the basic weight for flights, For flights At any moment The comprehensive priority score is a weighted fusion of time urgency, guarantee complexity, and basic weights, used to determine whether a flight needs to enter the critical coordination pool for key monitoring. The coefficients are updated regularly based on historical data to achieve adaptive optimization of priority evaluation; When flight time urgency is detected Exceeding the critical threshold The three-tiered judgment mechanism is activated: first, it checks whether the difference between the remaining support time and the planned takeoff time has entered the critical range [15 minutes, 45 minutes]; second, it assesses whether the support status meets the entry conditions; and finally, it confirms the results in conjunction with air traffic control constraints. After all passes, an entry command is generated. The system sends tiered warnings to designated support units and simultaneously activates the resource scheduling engine to calculate the optimal alternative and distribute it to the execution units. If a flight that has undergone manual intervention meets the rollout criteria, its priority is reassessed and added to the main sorting queue for recalculation. The real-time progress monitoring includes: The first step involves collecting multi-source data and constructing a distributed IoT sensor network. Trailers and refueling trucks are equipped with 5G-based positioning modules and status monitoring sensors, and data is transmitted via the MQTT protocol. Simultaneously, a video analytics system using an improved YOLOv8 algorithm identifies key operations such as boarding bridge docking and baggage handling. The advanced YOLOv8 algorithm uses airport cameras to identify key support processes in real time, including boarding bridge docking status, baggage cart loading / unloading actions, and refueling truck operation status. The visual recognition results are converted into support progress data, automatically updating flight support status, replacing manual reporting, and improving data accuracy and real-time performance. The second step is to construct the spatiotemporal state vector, using the flight number as a unique identifier, to build a multidimensional spatiotemporal state vector: ,in, To ensure progress, the report covers the completion status of 17 standard assurance tasks; The resource status component records the location and working status of the guaranteed resources; The table is divided into time components, including key time points planned / expected / actual. To constrain components, record flow control commands and meteorological restriction information; This is used as an event component to record historical anomalous events; The state vector is updated every 1 second, and the completion rate of the guaranteed node is calculated. With delay risk coefficient ; The third step is to design a dynamic threshold triggering mechanism and construct a three-layer cascaded triggering mechanism, including a basic threshold layer, a scene adaptation layer, and a history learning layer, to realize the dynamic generation of thresholds. The fourth step involves generating a reminder instruction. When the remaining coverage time for a flight is below the trigger threshold, the system identifies delayed coverage items and responsible entities, determines priorities based on delay risk, and generates a structured reminder instruction using a template engine. It sends data to the responsible unit through multiple channels, while recording complete operation logs to form a closed-loop tracking system.

7. The sorting method for airport flight coordination based on multi-source data as described in claim 1, characterized in that, In step S6, the optimization of trailer resource allocation includes: Based on the dynamic flight launch ranking results, an intelligent scheduling engine for trailer resources is constructed. By using IoT positioning technology, the location, operation status and remaining operation time of trailers are monitored in real time. Combined with the completion rate of flight support nodes and taxiing path planning requirements, a multi-objective optimization algorithm is used to construct a trailer-flight matching matrix. Multi-objective optimization algorithms include: constructing a weighted objective function. The expression is: ； In the formula, All of these are dynamically adjusted weighting coefficients. This represents the total task completion time. Total distance traveled. For energy consumption, This refers to the safety margin factor. An improved non-dominated sorting genetic algorithm, NSGA-II, is applied to generate a Pareto optimal solution set. The diversity and convergence of solutions are ensured through fast non-dominated sorting, crowding distance calculation, and elite retention strategies. Finally, a fuzzy comprehensive evaluation mechanism is adopted to select the best matching scheme from the Pareto front, achieving the optimal allocation of trailer resources and flight demand. A dynamic priority strategy is introduced to automatically allocate nearby trailer resources for critically delayed flights and VIP flights, simultaneously rehearse taxiing path conflict risks, and generate trailer scheduling routes without cross-interference. A dynamic priority strategy is introduced to monitor flight status in real time. When a flight is detected to be approaching a critical delay or is marked as a VIP flight, its trailer allocation priority is automatically increased, and the nearest available trailer is forcibly assigned to it. The trailer scheduling of other flights is also re-planned to ensure that important flights are given priority.

8. The sorting method for airport flight coordination based on multi-source data as described in claim 1, characterized in that, In step S6, the analysis of the causes of delay includes: Based on flight scheduling results and real-time support data stream, a dynamic tracing model for the root causes of delays is constructed. The construction process includes: first, establishing a time point deviation detection network to decompose flight operation trajectories and plans into event sequences and calculate deviation vectors; second, constructing a Bayesian network G(V,E) to identify causal dependencies; and finally, applying spatiotemporal sequence dimensionality reduction technology to form a panoramic view of delay evolution. A multi-level contribution metric framework is employed to identify key delay contributing factors and quantify their contribution weights. This framework utilizes an improved random forest algorithm to screen feature importance I(f). The Shapley method is introduced to decompose delay instance feature attribution. A hierarchical weight model is then constructed. ； In the formula, These are all weighting coefficients for each dimension. Reasons for delay The degree of time-related influence The posterior probability of cause C of delay. The importance of the characteristics of cause C for delay; Calculate the posterior probability and contribution weight of the root causes of delay using Bayesian inference. ; The association rule mining process employs an enhanced Apriori algorithm to extract high-frequency delay patterns; applies graph convolutional networks to model delay propagation; identifies key nodes based on a spatiotemporal attention mechanism; generates interpretable rule sets through decision tree paths, converting them into a visual source tracing report containing a heatmap of delay source distribution, a propagation link graph, a contribution ratio graph, and a trend prediction graph; and automatically pushes targeted optimization suggestions to responsible units.

9. The sorting method for airport flight coordination based on multi-source data as described in claim 1, characterized in that, In step S7, a dynamic re-entry mechanism is established for flights with resource updates or status changes. When a flight triggers a reordering condition due to updates to the support progress, adjustments to trailer scheduling, or the lifting of airspace restrictions, the priority coefficient is automatically updated based on real-time status parameters. The dynamic reentry mechanism comprises a three-layer trigger-evaluation-execution architecture: the trigger layer uses rule-based event listeners to capture real-time status change signals such as support progress updates, trailer scheduling adjustments, or airspace restriction removal; the evaluation layer uses a state difference detection algorithm to calculate the flight state vector. Compared with the state vector at the previous decision time Distance function When the distance exceeds the adaptive threshold When the reentrancy condition is met, the execution layer calls the priority recalculation function and performs queue reconstruction. Priority coefficients are automatically updated based on real-time status parameters, using a comprehensive scoring formula: ； In the formula, For dynamic time-varying weighting coefficients, For flights exist Feature values ​​of each dimension at any given time; The weight distribution is automatically adjusted according to the operational phase, weather conditions, and resource load, and an ε-greedy strategy is introduced to balance exploration and utilization. The dynamic weight allocation algorithm balances the original flight priorities by introducing a history-current state balance function. Adaptability to the new state The expression is: ； In the formula, It is a time decay function. The weight of the new state is gradually reduced as the waiting time increases, while maintaining the global sorting stability index. When a flight re-entry causes index fluctuations exceeding a preset threshold, a smoothing factor dynamic adjustment mechanism is triggered. This mechanism detects the fluctuation amplitude of the sorting stability index and dynamically adjusts the smoothing coefficient based on the degree of fluctuation, limiting the number of flights that change sorting in a single instance. This ensures a smooth transition in queue adjustments and avoids system oscillations caused by frequent re-sorting.

10. A sorting system derived from airport flight coordination by integrating multi-source data, characterized in that, The system implements the sorting method derived from airport flight coordination based on multi-source data as described in any one of claims 1-9, and the system includes: The multi-source data acquisition module is used to acquire data from multiple data sources related to flights. The multi-source data fusion module is used to fuse collected data from multiple sources to generate a five-dimensional dynamic situation matrix, which reflects the overall status of flights in real time. The module for determining the priority of manual sorting is used to determine whether to perform manual sorting priority processing based on the obtained comprehensive status; if yes, it will make optimization target decisions and intelligent sorting; otherwise, it will jump to the alternative sorting scheme display. The optimization target decision-making and intelligent sorting module is used to optimize target decisions and intelligent sorting based on the results of manual sorting and priority processing, by constructing an intelligent decision-making system for aircraft departure sorting. The alternative sorting scheme display module is used to provide multi-dimensional visual early warnings for real-time flights based on the real-time flight sequence optimization results, with flight bars including adjacent gate prompts and manual intervention functions. The early warning results are displayed as alternative sorting schemes in a visual sorting interface. The module for obtaining the ranking results of flight coordination is used to perform critical coordination pool processing, real-time support progress monitoring, trailer resource allocation optimization, and delay cause analysis on the obtained alternative ranking schemes to obtain the ranking results of flight coordination. The dynamic re-entry mechanism module is used to establish a dynamic re-entry mechanism for flights with resource updates or status changes based on the ranking results obtained from flight coordination, and to re-insert them into the global ranking queue; through a dynamic weight allocation algorithm, the original priority of the flight is balanced with the adaptability of the new state.

Images