Departure flight multi-target scheduling optimization method considering CTOT coincidence rate

Abstract

The embodiment of the invention discloses a departure flight multi-target scheduling optimization method considering a CTOT coincidence rate, and the method comprises the steps: collecting airport operation data, and building a scheduling sample library containing flight plan information, actual wheel chock push-out / withdrawal time, CTOT and other parameters; constructing a multi-objective optimization model fusing SOBT constraints and a COBT dynamic adjustment mechanism, and taking a departure normal rate, a CTOT coincidence rate and sliding time as joint optimization objectives; determining a multi-target weight by adopting an entropy weight method, and designing a hybrid simulated annealing algorithm fusing an elitism strategy and a self-adaptive weight mechanism to solve the model; and outputting the adjusted push-out moment plan according to an optimization result. The ground traffic scheduling and collaborative decision-making system is suitable for a large hub airport. The method aims to solve the problem of target conflict between the departure normal rate and the CTOT coincidence rate in the current airport departure scheduling process, and multi-target collaborative optimization of flight departure scheduling is realized by considering the deduction time control and taxiing conflict management.

Description

Technical Field

[0001] The present invention relates to the field of civil aviation air traffic control technology, and in particular to a multi-objective scheduling optimization method for departing flights that takes into account the CTOT (Calculated Take-Off Time) compliance rate. The method is suitable for flight push-out scheduling and flight release scheduling within the ground taxiing area of ​​a civil transport airport, and is mainly applied in the intersection of ground-air collaborative optimization and airspace capacity management. Background Art

[0002] As the core hub of the air transportation system, the ground operation efficiency of airports directly affects the punctuality of flights and the rational allocation of airspace resources. The International Civil Aviation Organization (ICAO) clearly stated in the "Global Air Navigation Plan" that airport operation efficiency should be evaluated with key performance indicators (KPIs), among which "departure normal rate" (KPI 01), "extra taxi-out time" (KPI02) and "CTOT compliance rate" (KPI 03) are the core indicators for measuring the efficiency of airport ground scheduling and air-ground coordination. How to improve the compliance of CTOT instructions while ensuring the normal departure of flights has become an important issue that needs to be solved in the current complex airport operation environment. It is also the further optimization direction of the civil aviation air traffic control system.

[0003] The departure regularity rate, typically measured as the proportion of flights whose actual rollout time (AOBT) deviates from the scheduled rollout time (SOBT) by no more than 15 minutes, is a primary indicator for evaluating ground scheduling timeliness and operational stability. The CTOT compliance rate, on the other hand, measures whether a flight departs within the time window allowed by its calculated takeoff time (CTOT) and is a key criterion for reflecting whether a flight is operating within the scheduled airspace capacity. Clearly, there is a significant conflict between the two objectives: to ensure the departure regularity rate, dispatchers must arrange flight rollouts as closely as possible to the original SOBT. However, if a flight fails to obtain an available taxi path or runway window at this time, taxiing congestion or conflicts may occur, impacting CTOT execution. Conversely, forcibly adjusting the rollout sequence or delaying taxiing to ensure CTOT compliance may result in irregular flight rollouts, impacting the overall timeliness of the flight's departure.

[0004] Currently, most airports still rely on fixed schedules within their dispatching systems, using rules based on the experience of experienced dispatchers or static priorities for push-back scheduling. These systems lack a systematic optimization mechanism for CTOT collaboration, making it difficult to dynamically balance conflicts between taxi paths between flights, airspace control slot execution requirements, and ground operation pressures. Furthermore, existing scheduling strategies lack the ability to model the linkage between SOBT and COBT, making it impossible to achieve comprehensive coordination between departure efficiency and airspace compliance during push-back taxiing. This results in unstable flight operations and low ground-air coordination, impacting overall airport operational efficiency. Summary of the Invention

[0005] An embodiment of the present invention provides a multi-objective scheduling optimization method for departing flights that takes into account the CTOT compliance rate. It can solve the target conflict problem between the departure normal rate and the CTOT compliance rate in the current airport departure scheduling process, and realize multi-objective collaborative optimization of flight departure scheduling by taking into account push-out time control and taxiing conflict management.

[0006] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0007] A multi-objective scheduling optimization method for departing flights taking into account CTOT compliance rate includes:

[0008] S1. Collect airport operation data and establish a scheduling sample library. The scheduling sample library includes: flight plan information, actual pushback / off-block time, and CTOT; specifically, basic flight information: flight number, scheduled pushback time (SOBT), calculated off-block time (COBT), actual off-block time (AOBT), calculated takeoff time (CTOT), actual takeoff time (ATOT); flight aircraft type, parking stand number, runway number, departure point direction, etc.; environmental status information: parking stand longitude and latitude coordinates, runway longitude and latitude coordinates, taxiway node coordinates, etc.

[0009] S2. Establish a multi-objective optimization model for departing flights that takes into account CTOT compliance. This multi-objective optimization model integrates the Scheduled Off-Block Time (SOBT) constraint and the Calculated Off-Block Time (COBT) dynamic adjustment mechanism. The multi-objective optimization model aims to maximize the departure regularity rate, maximize the CTOT compliance rate, and minimize the taxi time. The benchmark inputs for this multi-objective optimization model include the smooth taxi time.

[0010] S3. Determine and assign the weights of each objective function using the Entropy Weight Method (EWM);

[0011] S4. Run the multi-objective optimization model and adjust the launch time plan.

[0012] In this embodiment, in S1, the collection of airport operation data includes: collecting the full amount of flight operation data of the target airport within a specified period from the airport collaborative decision-making system (A-CDM) and the air traffic control automation system, wherein the types of the collected full amount of flight operation data include: basic flight information, taxi path information, ground resource and restriction information, and environmental status information; specifically, first, before flight scheduling modeling, the full amount of flight operation data of the target airport within a certain period is collected and pre-processed to form a high-quality scheduling input sample set. The collected data includes but is not limited to the following: 1. Basic flight information: flight number, scheduled out-of-block time (SOBT), calculated off-block time (COBT), actual off-block time (AOBT), calculated take-off time (CTOT), actual take-off time (ATOT); 2. Taxi path information: parking stand number, runway usage number, taxiway structure, and nodes along the taxi path; 3. Ground resources and constraints: taxi section occupancy time, pushback interval limit, taxi time, etc.; 4. Environmental status information: auxiliary fields such as flight type, airspace flow control classification, and whether it is a CTOT-controlled flight. The raw data is pre-processed for standardization, including field format conversion, unit unification, outlier removal, missing data completion, and taxi path mapping, ultimately constructing a standardized scheduling sample set for the model.

[0013] The surface structure data of the target airport is collected, and a directed graph model of the airport taxiing area is constructed. In addition, the airport surface structure data is collected, and a directed graph model of the airport taxiing area is constructed based on graph theory methods, where nodes represent parking stands, taxiway intersections, and runway entrances, and edges represent taxiway segments. Operational constraint parameters such as apron conflict intervals, wake turbulence intervals, departure direction restrictions, taxiing distances, etc. are integrated into the graph model. For example: The graph theory method is used to construct an airport surface operation model to collect airport surface operation data, including the latitude and longitude information, connection relationships, taxiing directions and distances of key nodes such as parking stands, runways, and taxiways, and to construct an airport surface taxiing network topology diagram. Nodes are used to represent parking stands, taxiway intersections, and runway entrances, and edges represent taxiing paths. A graph theory method is used to construct an airport surface operation model with a directed graph structure. Based on this graph model, relevant operational constraint information is integrated, including: apron pushback conflicts: the pushback time between adjacent parking stands must meet the minimum safe interval; wake turbulence interval constraints: the time interval required for takeoff between different aircraft types; departure direction interval: the minimum time interval between flights at the same departure point; the above constraints can be used as weights between nodes or attributes of edges in the graph structure, and are used for conflict detection and path optimization in subsequent model solving.

[0014] In this embodiment, in S1, the establishment of the scheduling sample library includes: flight operation data extracted and standardized from the collected full flight operation data, and the corresponding smooth taxi time for each runway-parking stand combination. The flight operation data includes: planned off-block time (SOBT), calculated off-block time (COBT), actual off-block time (AOBT), calculated takeoff time (CTOT), actual takeoff time (ATOT), flight aircraft type, parking stand location, runway in use, taxi time, and taxi path;

[0015] Standardization of raw flight operation data involves: data cleaning: removing unreasonable data such as time logic errors, missing fields, and taxi time outliers; unit unification: converting time fields to a standard format and matching location fields to node numbers in the graph model; and feature construction: extracting information such as the corresponding graph path, CTOT type indicator variables, and whether the flight is a controlled flight. Finally, a sample library is constructed for model training and scheduling optimization.

[0016] The smooth taxi time corresponding to each runway-parking stand combination includes: after data cleaning, grouping the remaining flights according to the departure stage, parking stand and runway usage to form runway-parking stand combinations; for each combination, the 10th percentile of the actual taxi time of all flights in the group is used as the corresponding smooth taxi time, wherein at least 10 flights in each combination have a smooth taxi time less than the smooth taxi time of the group.

[0017] In order to accurately reflect the regular time consumption during the taxiing phase of a flight and avoid outliers affecting the model performance, it is necessary to estimate the representative taxiing time for each parking stand-runway combination. In this scheme, "smooth taxiing time" is used as the input indicator for taxiing efficiency evaluation, which is defined as the time from the aircraft being pushed out of the parking stand to the time it arrives at the runway gate waiting in line, excluding the time waiting at the runway gate or release time. The specific steps are as follows: (1) Data cleaning: First, flights with de-icing procedures and helicopter flights are eliminated. Second, after calculating the actual taxiing time of the aircraft, flights with actual taxiing time less than 0 or greater than 2 hours are eliminated to ensure the validity of the data. (2) Grouping and clustering: Flights are grouped according to the parking stand and runway usage according to the departure phase, forming runway-parking stand combinations. For each airport, flights are grouped by combination with the same runway and parking stand. These runway-parking stand flight combinations follow similar taxiing paths and therefore have similar smooth taxiing times. (3) Calculation of Smooth Taxi Time: For each runway-stand combination, the smooth taxi time is calculated, which is the 10% quantile of the actual taxi time of all flights in the group. At the same time, to ensure that the taxi time of each runway-stand combination is representative, at least 10 flights in each combination have a taxi time shorter than the smooth taxi time of the group. The smooth taxi time is used as the benchmark input for flight taxi time estimation in the optimization model, used to measure the operational efficiency level between pushback and takeoff time, and provides an important reference in algorithm path calculation and taxi conflict detection.

[0018] In this embodiment, the actual push-out time (AOBT) of the flight is used as the main decision variable of the multi-objective optimization model; specifically, a multi-objective flight push-out scheduling optimization model for CTOT collaboration is constructed, aiming to achieve a dynamic balance between the following three objectives: maximizing the departure regularity rate: that is, launching flights within the SOBT ±15 minute window as much as possible; maximizing the CTOT compliance rate: that is, flights with CTOT restrictions should complete takeoff within the range of CTOT ±3 minutes; minimizing the taxiing time (smooth taxiing time): that is, reducing the ground taxiing time from push-out to take-off, and improving taxiing efficiency.

[0019] The model uses the actual push-out time (AOBT) as the primary decision variable and introduces the following key constraints: a SOBT reference window constraint sets a limit on the tolerance for flight push-out deviations; a COBT dynamic adjustment mechanism allows COBT adjustments under certain rules to meet CTOT constraints; a graphical taxi path feasibility constraint ensures that the selected path is conflict-free under current resource conditions; a CTOT indicator variable enables the CTOT window constraint only for controlled flights; resource occupancy and time interval constraints control key operational constraints such as the push-out interval and the release interval between adjacent flights; and taxi network topology modeling strengthens path connectivity assessment, node resource conflict resolution, and time sequence management. The model's objective function uses an entropy weighting method to weight and fuse, with weights reflecting the relative volatility and scheduling influence of each objective within the sample, achieving multi-objective coordination. The multi-objective optimization model includes: f=max(ω1f1+ω2f2-ω3f3), where ω1, ω2, and ω3 are weights corresponding to f1, f2, and f3, respectively, and f1, f2, and f3 are objective functions of departure normality, CTOT compliance rate, and taxiing time, respectively.

[0020] In this embodiment of the present invention, airport operation data is collected to establish a scheduling sample library containing flight schedule information, actual pushback / off-block times, and CTOT parameters. A multi-objective optimization model is constructed, integrating SOBT (Scheduled Off-Block Time) constraints and a COBT (Calculated Off-Block Time) dynamic adjustment mechanism, with departure on-time rate, CTOT compliance rate, and taxi time as joint optimization objectives. An entropy weight method is used to determine the weights of these multiple objectives, and a hybrid simulated annealing (HSA) algorithm, integrating an elite retention strategy with an adaptive weighting mechanism, is designed to solve the model. An adjusted pushback schedule is output based on the optimization results. This resolves the conflict between departure on-time rate and CTOT compliance in current airport departure scheduling processes. By balancing pushback time control and taxi conflict management, multi-objective collaborative optimization of flight departure scheduling is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0022] Figure 1 The smooth taxiing time algorithm process provided by the embodiment of the present invention;

[0023] Figure 2 A method for generating a feasible solution provided by an embodiment of the present invention;

[0024] Figure 3 The simulated annealing algorithm solution process provided by the embodiment of the present invention;

[0025] Figure 4 The effect diagram of the departure regularity rate before and after optimization of Lukou Airport in each period of the day provided by the embodiment of the present invention;

[0026] Figure 5 This is a diagram of the CTOT compliance rate before and after optimization for Lukou Airport during each period of the day, provided by an embodiment of the present invention;

[0027] Figure 6 Density map of Lukou Airport's monthly taxi time before and after optimization provided by an embodiment of the present invention;

[0028] Figure 7 A schematic diagram of the direction of an airport departure point provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0029] To enable those skilled in the art to better understand the technical solutions of the present invention, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments of the present invention will be described in detail below, with examples of the embodiments illustrated in the accompanying drawings. Throughout, identical or similar reference numerals represent identical or similar elements or elements having identical or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and intended only to explain the present invention and are not to be construed as limiting the present invention. Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a," "an," "said," and "the" used herein may also include the plural forms. It should be further understood that the term "comprising" as used in the description of the present invention refers to the presence of the stated features, integers, steps, operations, elements, and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when an element is referred to as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or intervening elements may be present. Furthermore, "connected" or "coupled" as used herein may include wireless connections or couplings. The term "and / or" as used herein includes any and all combinations of one or more associated listed items. It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art, and will not be interpreted in an idealized or overly formal sense unless defined as such herein.

[0030] This embodiment's design philosophy is to address conflicting objectives in airport ground scheduling by synergistically considering flight departure on-time rate, CTOT compliance, and taxi time. This approach is suitable for hub airports equipped with a coordinated departure and departure (CDM) system and taxi scheduling module, and is particularly well-suited for medium- and large-scale airports with dense operations, strict CTOT restrictions, and intense competition for taxi area resources.

[0031] The design objectives of this embodiment are to: improve the execution compliance of flight CTOT instructions while ensuring the regular flight departure rate; effectively control the flight's extra taxi-out time and improve airport taxiing efficiency; realize intelligent optimization of the coordinated scheduling mechanism between flight SOBT, AOBT and COBT; and build a multi-objective scheduling model for complex airport operating environments to provide decision support for airport operation management.

[0032] The embodiment of the present invention provides a multi-objective scheduling optimization method for departing flights taking into account the CTOT compliance rate, such as Figure 1-3 Shown, including:

[0033] S1. Collect airport operation data and establish a scheduling sample library, which includes flight plan information, actual push-out / off-block times, and CTOT;

[0034] S2. Establishing a multi-objective optimization model for departing flights that takes into account the CTOT compliance rate, wherein the multi-objective optimization model has as its optimization objectives maximizing the departure on-time rate, maximizing the CTOT compliance rate, and minimizing the taxiing time, and the benchmark input of the multi-objective optimization model includes the unimpeded taxiing time;

[0035] S3. Determine and assign the weights of each objective function using the Entropy Weight Method (EWM);

[0036] S4. Run the multi-objective optimization model and adjust the launch time plan.

[0037] In S1, the collection of airport operation data includes: collecting all flight operation data of the target airport within a specified period, wherein the types of the collected full flight operation data include: basic flight information, taxi path information, ground resource and restriction information, and environmental status information; collecting the surface structure data of the target airport, and constructing a directed graph model of the airport taxi area.

[0038] The establishment of the scheduling sample library includes: extracting and standardizing flight operation data from the collected full flight operation data, including: planned off-block time (SOBT), calculated off-block time (COBT), actual off-block time (AOBT), calculated take-off time (CTOT), actual take-off time (ATOT), flight aircraft type, parking stand location, used runway, taxi time and taxi path; and the clear taxi time corresponding to each runway-parking stand combination.

[0039] Among them, such as Figure 1 The clear taxi time calculation method shown in the figure shows the clear taxi time corresponding to each runway-parking stand combination. The method includes: after data cleaning, the remaining flights are grouped according to the departure stage, the parking stand and the runway usage to form runway-parking stand combinations; for each combination, the 10% quantile of the actual taxi time of all flights in the group is used as the corresponding clear taxi time, wherein at least 10 flights in each combination have a clear taxi time less than the group's clear taxi time.

[0040] The actual push-out time (AOBT) of the flight is used as the main decision variable of the multi-objective optimization model;

[0041] The multi-objective optimization model includes: f=max(ω1f1+ω2f2-ω3f3), where ω1, ω2, and ω3 are weights corresponding to f1, f2, and f3, respectively, and f1, f2, and f3 are objective functions of departure normality, CTOT compliance rate, and taxiing time, respectively.

[0042] Specifically, the departure normal rate, CTOT compliance rate and taxiing time are taken as joint optimization objectives. The departure normal rate objective function is based on:

[0043] |AOBT-SOBT|≤15 minutes definition;

[0044] The CTOT compliance rate is defined based on a window of |ATOT-CTOT| ≤ 3 minutes; the taxi time is ATOT-AOBT.

[0045] To achieve the three-objective collaborative optimization. The variables are defined as shown in Table 1 below:

[0046] Table 1

[0047]

[0048]

[0049] The objective function is as follows:

[0050] ① Maximize the departure normal rate

[0051] The departure regularity rate is the percentage of regular departure flights at an airport to the total number of departure flights at the airport, while the criterion for judging punctuality is if the difference between the actual off-block time (AOBT) and the scheduled off-block time (SOBT) is within the allowed time range (|AOBT-SOBT|≤15 minutes).

[0052] The normal departure rate is:

[0053] Among them, f1 represents the normal departure rate, and the objective function of the normal departure rate is: Push-out time of aircraft i at parking stand g (actual off-block time AOBT) The planned off-block time (SOBT) δ(*) of the i-th aircraft at parking stand g is an indicator function that takes the value 1 when the condition is true and 0 otherwise;

[0054] ② Maximize CTOT compliance rate

[0055] Since not all departing flights include CTOT, a CTOT indicator variable is set to identify which flights have CTOT constraints.

[0056]

[0057] Among them, if the flight has CTOT, then If the flight does not have CTOT,

[0058] The CTOT compliance rate measures whether the flight's actual take-off time (ATOT) complies with the CTOT constraint, meaning whether the aircraft can complete take-off within the specified time window. Specifically, CTOT slot tolerances are divided into two categories: first-class and second-class. Capacity-based traffic management measures fall within the first category, with a tolerance range of (-5, +10) minutes; interval-based traffic management measures fall within the second category, with a tolerance range of (-3, +3) minutes. Therefore, a tolerance range of (-3, +3) minutes is chosen in this paper.

[0059] The maximum CTOT compliance rate is defined as:

[0060]

[0061] Among them, δ(*) is the indicator function, when The value is 1 if it is minutes, otherwise it is 0.

[0062] in, represents the actual take-off time of the i-th aircraft at the r-th runway, It indicates the take-off time of the i-th aircraft at the r-th runway, represents an indicator variable used to identify flights containing CTOT;

[0063] ③ Minimize taxiing time

[0064] The taxi time in this paper is the actual take-off time of the aircraft minus the pushback time of the aircraft. Since the taxi path time in this paper is replaced by the smooth taxi time, minimizing the taxi time means minimizing the extra taxi time:

[0065]

[0066] A multi-objective function is established:

[0067]

[0068] Among them, ω1, ω2, and ω3 use the Entropy Weight Method (EWM) to calculate the weights of each objective function.

[0069] S3 includes: determining the weight of each objective function by an entropy weight method according to the constraint conditions, wherein the constraint conditions include: parking stand conflict constraint, runway release interval constraint and CTOT time window constraint;

[0070] The parking stand conflict constraint is a minimum pushback time interval requirement: the minimum pushback time of adjacent impacted stands is limited to a certain time:

[0071]

[0072] To ensure that departing flights meet the departure direction spacing and aircraft type spacing requirements when entering the runway, we define the entry time of aircraft i on runway r as The release interval is The runway clearance interval constraint is:

[0073] Aircraft enters the runway at:

[0074] The release interval is determined by the departure direction interval and the aircraft type interval. The final release interval constraint is:

[0075] The actual take-off time is the sum of the runway entry time and the runway occupation time:

[0076]

[0077] The CTOT time window constraint is:

[0078]

[0079] in, represents a set of aircraft numbers, i and j represent the aircraft numbers in the set, represents the parking stand set, g represents the parking stand number, represents the time when the i-th aircraft enters runway r, represents the push-out time of the i-th aircraft, U gr represents the clear taxiing time from parking stand g to runway r, It indicates the take-off time of the i-th aircraft at the r-th runway, represents the actual take-off time of the i-th aircraft at the r-th runway, represents an indicator variable used to identify flights containing CTOT, A comprehensive factor representing runway occupancy time,

[0080] Indicates runway clearance interval.

[0081] Specifically, The runway occupancy time is a combination of multiple factors, such as wind direction and speed, runway wetness, load, crew operation, etc. The value is differentiated by aircraft type, which includes at least light, medium and heavy aircraft. For example, the runway occupancy time requirements for each aircraft type are shown in Table 2.

[0082] Table 2

[0083] Approach Exit mid-range machine 50s 50s Heavy Machine 60s 60s

[0084] The time when the aircraft enters the runway is:

[0085]

[0086] The clearance interval is determined by the departure direction interval and the aircraft type interval. For example, due to the differences between different airports, This can be directly interpreted as the runway clearance interval requirements. The following content examples (departure direction interval and aircraft type interval) are based on the Lukou Airport in a provincial capital city as an example:

[0087] (1) Departure interval

[0088] like Figure 7 As shown, departing aircraft passing the same position reporting point (HFE, ESBAG, OF, CJ) are considered to be heading in the same direction, and the interval in the same direction is 3 minutes; and the directions to Zhengzhou, Wuhan, and Nanchang after HFE are 6 minutes; different directions meet the wake turbulence separation requirements (the minimum is generally assumed to be 2 minutes), such as Figure 7 shown.

[0089]

[0090] (2) Model interval Its value depends on the aircraft type combination of the preceding and following flights

[0091] The release interval requirements for each aircraft type are shown in Table 3:

[0092] Table 3

[0093]

[0094] The release interval is:

[0095]

[0096] The final release interval constraint is:

[0097]

[0098] In order to reasonably allocate the weights of each optimization objective, this embodiment adopts the Entropy Weight Method to determine the weight coefficients of each objective function. The Entropy Weight Method is an objective weighting method that allocates weights based on the degree of variation of each indicator in the sample data. The steps are as follows:

[0099] (1) Standardize the objective function values ​​and normalize the values ​​of each objective function to avoid the influence of dimension differences on the calculation results;

[0100] (2) Calculate the proportion of each objective function in each sample;

[0101] (3) According to the information entropy formula, the entropy value of each objective function is calculated. The smaller the entropy value, the greater the variability of the objective.

[0102] (4) Redundancy (i.e., information gain) is inferred from the entropy value, and the weight of each target is calculated accordingly. The greater the redundancy, the more information the target provides, and the higher the optimization priority should be.

[0103] This method ensures objectivity in weight allocation, giving higher weights to objectives with a greater impact on airport operations, thereby enhancing the model's adaptability and fairness. Example weights are: ω1 = 0.31 (on-time departure rate), ω2 = 0.44 (CTOT compliance rate), and ω3 = 0.25 (taxi time).

[0104] In order to efficiently solve the above multi-objective scheduling optimization problem, the present invention adopts a multi-objective optimization algorithm framework based on heuristic search, including:

[0105] Multi-objective evolutionary algorithms (such as NSGA-II or MOEA / D): obtain a set of non-dominated solutions in the solution space through a multi-objective Pareto optimization strategy;

[0106] Hybrid improved algorithm: It can combine local search strategies such as particle swarm optimization (MOPSO) and simulated annealing (HSA) to improve the convergence and diversity of solutions;

[0107] Fitness function construction: combining the standardized scores of the three objectives and the dynamic congestion penalty factor;

[0108] Population initialization mechanism: The initial population is constructed based on the SOBT priority rule to ensure the feasibility of the initial solution;

[0109] Elite retention and congestion control mechanism: Improves the algorithm's local search stability in high-density scheduling intervals;

[0110] Adaptive target weight adjustment mechanism: automatically adjusts the bias strategy according to the target convergence trend during the evolution process.

[0111] The optimization framework can ultimately output a set of optimal or suboptimal scheduling solutions that meet multi-objective coordination for selection in actual deployment systems or reference in multi-scenario analysis.

[0112] A hybrid metaheuristic framework combining simulated annealing (SA) with a local search strategy is used to solve the multi-objective flight rollback scheduling problem. By balancing global exploration with local exploitation, the algorithm efficiently converges to a high-quality Pareto optimal solution under complex operational constraints. The algorithm design primarily involves data preprocessing, feasible solution generation, and conflict resolution. S4 includes:

[0113] (1) Preprocessing: The parameter set of the multi-objective optimization model is:

[0114]

[0115] In the computational preprocessing phase, data cleaning and feature extraction are performed: outlier removal: gliding time is deleted Grouping and clustering: Divide the flight set by the parking position-runway combination and calculate the smooth taxiing time U for each group gr ..

[0116] (2) This embodiment proposes a scheduling strategy based on the combination of SOBT anchoring and COBT window elastic adjustment to enhance the adaptability and execution flexibility of flight rollback scheduling, thereby completing the establishment of a dynamic time adjustment mechanism, which is specifically reflected in the generation of feasible solutions, such as:

[0117] For flight i, its feasible departure time interval Defined as:

[0118]

[0119] Flights with CTOT: The time window is strictly limited to 10 minutes before and after COBT.

[0120] Flights without CTOT: The time window is extended to 15 minutes after the scheduled departure time.

[0121] Further information on the parking space occupancy schedule D g Perform conflict detection: eliminate unavailable periods.

[0122]

[0123] D g Indicates the parking stand occupancy schedule, the occupancy time interval on parking stand g.

[0124] The feasible departure time of the parking stand is added to the smooth taxiing time to obtain the initial arrival time interval

[0125] For runway opening r, the conflict-free period is generated according to the following rules: (1) Time window division: according to the runway release interval Will Discretize into time slots. (2) Conflict detection: Eliminate the time and space overlap intervals with the scheduled flights. (3) Priority allocation: Prioritize the reallocation of time windows for flights CTOT in the conflicting time slots. Among them, time slots are due to some intervals, resulting in the original continuous time being occupied due to the required interval time. It can only be used as the initial arrival time interval in other time slots.

[0126] Determine the actual departure time of the gate flight: Reverse calculation: Determine the gate departure time from the actual runway entrance arrival time. Constraint check: Ensure that the departure time is still within the initial feasible time window.

[0127] Update the parking stand occupancy table: Primary parking stand: record the actual departure time for subsequent flight conflict detection. Prohibit pushback: add parking stand time interval constraints in the adjacent stand schedule. Update the parking stand occupancy table The specific algorithm solution is as follows:

[0128] a) Initialization

[0129] The initial solution X is generated by greedy heuristics, with the initial temperature T0 = 1000, the cooling rate α = 0.95, the maximum number of iterations 500, the parking spaces allocated according to the SOBT sorting, and the feasible time window filled greedily.

[0130] b) Annealing cycle

[0131] For each iteration,

[0132] (1) Neighborhood generation: Generate the neighborhood solution X′ of the current solution;

[0133] (2) Target evaluation: F(X′) = ω1·f1+ω2·f2-ω3·f3;

[0134] (3) Acceptance criterion: If F(X′)>F(X t ), accept X t+1 =X′; otherwise, with probability accept;

[0135] (4) Elite archive update: retain the top 10% optimal solutions to accelerate convergence;

[0136] Temperature update: T t+1 =α·T t Based on the above design, it is implemented as a three-level coordinated optimization strategy in actual application. In addition to the above-mentioned SOBT-based benchmark anchoring, it also includes the elastic correction of AOBT (Actual Off-Block Time) and the fine matching of COBT. For example: AOBT is the actual off-block time, and it can be elastically changed according to +~15 minutes of SOBT or 10 minutes before and after COBT (i.e., elastic correction of AOBT); for flights including CTOT: the time window is strictly limited to 10 minutes before and after COBT (i.e., fine matching of COBT), and AOBT (actual off-block time) is executed within this time window.

[0137] Empirical verification and application output stage: This embodiment conducted a large-scale verification test based on the real flight data of Lukou International Airport in a provincial capital city, and built a scheduling simulation environment based on 7,658 departing flights in June 2023. By comparing the scheduling indicators before and after optimization, the applicability and superiority of the method of the present invention in actual combat were verified: Figure 4-6 As shown, taxiing efficiency has improved: the average taxiing time has been reduced by 6.4%, the extreme long-tail taxiing time has been reduced by 58%, and the additional taxiing time of key combination flights has been reduced by more than 82%; the scheduling coordination capability has been improved: the departure normal rate has increased by 12.1%, the CTOT compliance rate has increased by 33.7%, and the consistency of collaborative scheduling has been enhanced; the system deployment feasibility is high: the scheduling results can be used to build an artificial decision-making support system or an intelligent scheduling platform to support the collaborative operation of the airport, air traffic control and airlines.

[0138] Compared with the prior art, this embodiment has the following significant advantages and beneficial effects: Multi-objective coordinated optimization: Different from the traditional single-objective or static rule scheduling method, the present invention establishes a joint optimization objective function with departure normality rate, CTOT compliance rate and taxiing efficiency as the core, to achieve coordinated control of multiple indicators in airport ground scheduling. Dynamic time adjustment mechanism: The present invention proposes a scheduling strategy based on the combination of SOBT anchoring and COBT window elastic adjustment to enhance the adaptability and execution flexibility of flight launch scheduling. Strong applicability and high promotion value: The present invention is applicable to hub airports under a variety of different operating environments, and is particularly suitable for deployment in the context of high-density flight operations. It can be widely used in ground taxiing scheduling systems, intelligent tower decision support systems and air-ground collaborative release platforms, and has good engineering implementation potential and practical application value.

[0139] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited to this. Any changes or replacements that can be easily thought of by any technician familiar with this technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A multi-objective scheduling optimization method for departing flights taking into account CTOT compliance rate, characterized by: include: S1. Collect airport operation data and establish a scheduling sample library, which includes flight plan information, actual push-out / off-block times, and CTOT; S2. Establishing a multi-objective optimization model for departing flights that takes into account the CTOT compliance rate, wherein the multi-objective optimization model has as its optimization objectives maximizing the departure on-time rate, maximizing the CTOT compliance rate, and minimizing the taxiing time, and the benchmark input of the multi-objective optimization model includes the unimpeded taxiing time; S3. Determine and distribute the weights of each objective function using the entropy weight method (EWM); S4. Run the multi-objective optimization model and adjust the launch time plan.

2. The method according to claim 1, characterized in that In S1, the collection of airport operation data includes: Collect all flight operation data of the target airport within a specified period. The types of flight operation data collected include basic flight information, taxi path information, ground resource and restriction information, and environmental status information. The surface structure data of the target airport is collected, and a directed graph model of the airport taxiing area is constructed.

3. The method according to claim 2, characterized in that In S1, the establishment of a scheduling sample library includes: Flight operation data extracted and standardized from the collected full flight operation data includes: planned off-block time (SOBT), calculated off-block time (COBT), actual off-block time (AOBT), calculated take-off time (CTOT), actual take-off time (ATOT), aircraft type, parking stand location, runway used, taxi time, and taxi path; and, the clear taxi time corresponding to each runway-stand combination.

4. The method according to claim 3, characterized in that The clear taxi time corresponding to each runway-parking stand combination includes: After data cleaning, the remaining flights are grouped according to the departure stage, parking positions and runway usage to form runway-parking position combinations; For each combination, the 10% quantile of the actual taxi time of all flights in the group is taken as the corresponding smooth taxi time, where at least 10 flights in each combination have a taxi time shorter than the smooth taxi time of the group.

5. The method according to claim 1, wherein The actual push-out time (AOBT) of the flight is used as the main decision variable of the multi-objective optimization model; The multi-objective optimization model includes: f=max(ω1f1+ω2f2-ω3f3), where ω1, ω2, and ω3 are weights corresponding to f1, f2, and f3, respectively, and f1, f2, and f3 are objective functions of departure normality, CTOT compliance rate, and taxiing time, respectively.

6. The method according to claim 5, characterized in that The objective function of the departure normal rate is: Among them, f1 represents the departure regularity rate, N represents the total number of flights, P i g represents the push-out time of the i-th aircraft at the g-th parking stand, represents the planned off-block time of aircraft i at parking stand g, δ(*) is an indicator function that takes the value 1 when the condition is true and 0 otherwise; The objective function of CTOT compliance rate is: in, represents the actual take-off time of the i-th aircraft at the r-th runway, It indicates the take-off time of the i-th aircraft at the r-th runway, represents an indicator variable used to identify flights containing CTOT; The objective function of the sliding time is:

7. The method according to claim 5, characterized in that Included in S3: Determine the weight of each objective function using the entropy weight method according to the constraints, wherein the constraints include: parking stand conflict constraint, runway release interval constraint and CTOT time window constraint; The parking space conflict constraints are: The runway clearance interval constraint is: The CTOT time window constraint is: in, represents a set of aircraft numbers, i and j represent the aircraft numbers in the set, represents the parking stand set, g represents the parking stand number, represents the time when the i-th aircraft enters runway r, P i g represents the push-out time of the i-th aircraft, u gr represents the clear taxiing time from parking stand g to runway r, It indicates the take-off time of the i-th aircraft at the r-th runway, represents the actual take-off time of the i-th aircraft at the r-th runway, represents an indicator variable used to identify flights containing CTOT, A comprehensive factor representing runway occupancy time, Indicates runway clearance interval.

8. The method according to claim 1 or 7, characterized in that Included in S4: The parameter set input into the multi-objective optimization model is: Calculate the feasible departure time of the parking stand and add the smooth taxiing time to obtain the initial arrival time interval Generate a conflict-free period for each runway entrance and determine the actual departure time of the flight occupying the gate; Updated parking slot occupancy schedule.

9. The method according to claim 8, characterized in that Generating a conflict-free period for each runway intersection includes: Runway clearance interval Will Discretization into time slots; Eliminate time and space overlaps with scheduled flights; Prioritize reallocation of time windows for CTOT flights within conflicting time slots.

Images