Airport bus network optimization method and system integrating air traffic information

Abstract

The application discloses an airport bus network optimization method and system integrating air traffic control information, and the method comprises the following steps: acquiring the geographical range of the airport bus operation area, passenger reservation demand and passenger online car-hailing order data generated by the influence of flight dynamic changes according to the air traffic control information; determining all airport bus station positions in the airport bus operation area; establishing a target function according to passenger travel time cost and airport bus operation cost, and establishing constraint conditions according to the maximum number of passengers served by the airport bus starting station, the service level of the airport bus service, the lower limit and the upper limit of the time window in which the passenger expects the airport bus to arrive; and solving a reservation response bus scheduling model to obtain an airport bus scheduling scheme. Under the condition of flight changes, the number of passengers taking the airport bus and their destinations can be known in advance, the operation strategy can be quickly adjusted, and the vehicle scheduling can be optimized, so that the waiting time and in-vehicle time of passengers can be effectively saved, and the passenger satisfaction under the condition of flight changes can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of airport bus resource scheduling, and in particular to an airport bus network optimization method and system integrating air traffic control information. Background Art

[0002] In modern air transportation, airport shuttles play a vital role, transporting passengers from airports to different parts of the city. This greatly facilitates travel and effectively promotes seamless transportation between airports and cities. However, the current operation model of airport shuttles has exposed some problems that need to be addressed.

[0003] First, flight conditions are dynamic, with frequent delays, cancellations, and other changes. This makes it difficult for airport shuttles to accurately predict passenger flow and distribution at different times. Traditional airport shuttle operations, with their fixed schedules and routes, are rigid and inflexible in the face of these dynamics. As a result, buses are often left idle during certain times of the day, while at other times, they face a shortage of capacity.

[0004] Secondly, the existing airport shuttle bus station layout is mostly based on past experience, without fully considering the dynamic changes in passengers' actual travel needs over time and the ever-evolving urban traffic conditions. This results in the inability to effectively meet passenger connection needs in some areas, or the operational efficiency of airport shuttles on urban roads is significantly reduced due to unreasonable station layout.

[0005] Furthermore, current airport shuttle bus operations generally utilize a fixed schedule with fixed departure intervals and stops. This model struggles to quickly respond to passengers' individual travel needs during off-peak hours. Passengers often face lengthy wait times during transfers, significantly impacting their travel experience. The drawbacks of this traditional operating model become particularly pronounced during flight schedule changes, making it difficult to promptly address unexpected passenger needs. Passengers during off-peak hours may face lengthy wait times or experience increased travel time due to poor route planning, severely impacting their travel experience.

[0006] Therefore, those skilled in the art are in urgent need of developing a new technical solution to solve the above problems. Summary of the Invention

[0007] In order to overcome the problems existing in the related art, the present invention discloses an airport bus network optimization method and system integrating air traffic control information.

[0008] According to a first aspect of the disclosed embodiments of the present invention, a method for optimizing an airport bus network by integrating air traffic control information is provided, the method comprising:

[0009] Obtain the geographical scope of the airport shuttle operation area, passenger reservation demands caused by flight dynamics, and passenger online ride-hailing order data based on air traffic control information. The passenger reservation demands include: number of passengers, reservation time window, and destination;

[0010] Performing grid processing on the operation area to divide the operation area into an airport bus operation area network consisting of a plurality of grid units;

[0011] Counting the passenger flow demand of each grid cell based on the distribution of the drop-off locations in the passenger online ride-hailing order data, and screening out high-frequency demand grid cells whose passenger flow demand is higher than a preset demand threshold;

[0012] Determining the location of the airport bus stop within the operating area according to the high-frequency demand grid unit;

[0013] Constructing an airport bus scheduling optimization model, wherein the objective function of the airport bus scheduling optimization model is the sum of minimized passenger travel time cost and minimized bus operating cost, and the constraints of the airport bus scheduling optimization model are established based on the maximum number of service passengers at the airport bus starting station, the service level of the airport bus schedule, the lower limit of the time window when passengers expect the airport bus to arrive, and the upper limit of the time window when passengers expect the airport bus to arrive;

[0014] A multi-layer coding optimization algorithm is used to solve the airport bus scheduling optimization model, and an airport bus operation route set, a departure schedule, and a passenger allocation plan are output to optimize the airport bus operation area network.

[0015] Optionally, determining the location of the airport bus stop within the operating area according to the high-frequency demand grid unit includes:

[0016] Merge adjacent high-frequency demand grid cells to form a first passenger flow hotspot, and then perform a secondary merger on every two first passenger flow hotspots whose centroid distance is less than or equal to a preset distance threshold;

[0017] The first passenger flow hot zone after the second merger is divided and adjusted so that the horizontal span distance and the vertical span distance of the adjusted hot zone do not exceed the preset maximum service scale, thereby forming a second passenger flow hot zone;

[0018] Each high-frequency demand grid cell in the second passenger flow hot zone is scored according to the number of adjacent grids of each high-frequency demand grid cell in the second passenger flow hot zone, and the high-frequency demand grid cell center with the highest score result is selected as the airport bus stop location.

[0019] Optionally, the counting of passenger flow demand of each grid cell based on the distribution of drop-off locations in the passenger online ride-hailing order data, and screening out high-frequency demand grid cells whose passenger flow demand is higher than a preset demand threshold, includes:

[0020] Perform a grid index mapping operation on the drop-off locations in the passenger's online ride-hailing order data to determine the grid cells corresponding to each drop-off location ;

[0021] Count the total number of passenger drop-off events in the grid unit within a preset time period as the passenger flow demand of the grid unit ;

[0022] Passenger flow demand Higher than the preset demand threshold Grid cells Determined as a high-frequency demand grid unit ;

[0023] Get the high-frequency demand grid cell set as .

[0024] Optionally, dividing and adjusting the first passenger flow hot zone after the second merger so that both the horizontal span distance and the vertical span distance of the adjusted hot zone do not exceed a preset maximum service scale to form a second passenger flow hot zone includes:

[0025] Calculate the maximum horizontal span distance of each first passenger flow hot zone and maximum vertical span distance ;

[0026] If the first passenger flow hot zone Maximum horizontal span distance Greater than , then the first passenger flow hot zone is horizontally Perform the segmentation, where The preset maximum service scale;

[0027] If the first passenger flow hot zone Maximum vertical span distance Greater than , then the first passenger flow hot zone is vertically Perform segmentation;

[0028] If the first passenger flow hot zone The maximum vertical span distance and the maximum horizontal span distance are greater than , then the first passenger flow hot zone is Perform segmentation;

[0029] The splitting position is determined based on the principle of minimizing the absolute value of the difference between the sum of passenger flow demands of all grid cells in the two hot zones formed after the splitting.

[0030] The first passenger flow hot zone is divided according to the segmentation direction and segmentation position. Split until the maximum horizontal span distance of all hot zones after splitting and maximum vertical span distance All less than , get the second passenger flow hot zone ;

[0031] The second passenger flow hot zone set within the geographical area where the airport bus operates is recorded as .

[0032] Optionally, the method for scoring each high-frequency demand grid cell in the second passenger flow hot zone is: ,in, For the scoring results, The second passenger flow hot zone The number of adjacent grids of a high-frequency demand grid cell ( ), is the passenger flow demand of the second passenger flow hot zone, and is the weight coefficient, Represents each grid cell, Represents a high-frequency demand grid cell.

[0033] Optionally, the objective function of the airport bus scheduling optimization model is expressed as: ,in, is the objective function, The travel time cost of passengers, For the operating costs of the airport shuttle, and is the weight coefficient, and the calculation formula for the passenger travel time cost is: ,in, , , , The time required for passengers to walk from the airport to the airport bus stop; The time the passenger is on board, To deviate from the time cost of service, Indicates passenger reservation demand With the airport bus If the passenger's reservation requirements By Airport Bus Providing services ,otherwise , is the distance between the passenger's airport and the airport bus stop, represents the average speed at which passengers walk to the airport bus stop. Indicates airport bus The starting point location and airport bus stop location The distance between Indicates airport bus From site location Drive to the station location The average running speed, Airport shuttle for passengers Arrival at the site The lower limit of the time window, The location of the airport bus stop that passengers expect to arrive at The upper limit of the time window, For airport buses Passengers arrive at the station location The calculation formula of the airport bus operating cost is: ,in, , , , , is the fixed cost of the vehicle, For variable operating costs, The penalty cost for the airport bus to deviate from the passenger's expected arrival time is: The penalty cost for the system rejecting the order, Indicates airport bus Whether to go from the starting station location to the station location If so, , if otherwise , is the depreciation factor for each operation of the airport bus, is the fuel cost per mile for the airport shuttle vehicle, The unit mileage cost for airport shuttle drivers, is the penalty coefficient for the actual arrival time of the airport bus being earlier than the passenger’s expected arrival time at the station, is the penalty coefficient for the actual arrival time of the airport bus being later than the passenger’s expected arrival time at the station. The penalty cost coefficient generated by the system rejecting the order, Gathering for the airport bus, Gather at the airport bus stop location.

[0034] Optionally, the airport bus network optimization method integrating air traffic control information is characterized in that the process of establishing the constraint conditions includes:

[0035] The first constraint is established based on the maximum number of service passengers at the starting stop of the airport bus: ,in, The maximum number of passengers served at the starting point of the airport bus. Indicates passenger reservation demand With the airport bus If the passenger's reservation requirements By Airport Bus Providing services ,otherwise , Collect reservation requirements for passengers, Gather for the airport bus;

[0036] The second constraint is established based on the service level of the airport bus service: ,in, For airport buses Minimum service level, For airport buses Maximum service level, Indicates airport bus Whether to go from the starting station location to the station location If so, , if otherwise , Gather for the location of the airport bus stop;

[0037] The third constraint is established based on the lower limit of the time window when passengers expect the airport bus to arrive and the upper limit of the time window when passengers expect the airport bus to arrive: ,in, The location of the airport bus stop that passengers expect to arrive at The lower limit of the time window, The location of the airport bus stop that passengers expect to arrive at The upper limit of the time window.

[0038] Optionally, the multi-layer coding optimization algorithm is used to solve the airport bus scheduling optimization model, including:

[0039] Gather the airport bus Defined as the first level encoding, where For airport buses The vehicle number of the shift;

[0040] Sequence the service appointment request is defined as the second level encoding, where For airport buses service appointment requests;

[0041] Departure time sequence Defined as the third level encoding, where For airport buses The vehicle number of the shift The corresponding departure time from the starting station;

[0042] The initial feasible solution is designed and generated using a three-layer coding method. .

[0043] According to a second aspect of the disclosed embodiments of the present invention, there is provided an airport bus network optimization system integrating air traffic control information, the system comprising:

[0044] A data acquisition module, which obtains the geographical scope of the airport shuttle operation area, passenger reservation demands caused by flight dynamics, and passenger online ride-hailing order data based on air traffic control information. The passenger reservation demands include the number of passengers, reservation time window, and destination.

[0045] a rasterization processing module connected to the data acquisition module, performing rasterization processing on the operation area so as to divide the operation area into an airport bus operation area network consisting of a plurality of raster units;

[0046] a high-frequency grid determination module, connected to the grid processing module, which calculates the passenger flow demand of each grid cell based on the distribution of the drop-off locations in the passenger online ride-hailing order data, and screens out high-frequency demand grid cells whose passenger flow demand exceeds a preset demand threshold;

[0047] a station location determination module, connected to the high-frequency grid determination module, and determining the location of the airport bus station within the operating area according to the high-frequency demand grid unit;

[0048] an optimization model construction module, connected to the station location determination module, for constructing an airport bus scheduling optimization model, wherein the objective function of the airport bus scheduling optimization model is the sum of minimized passenger travel time cost and minimized bus operating cost, and the constraints of the airport bus scheduling optimization model are established based on the maximum number of service passengers at the airport bus starting station, the service level of the airport bus schedule, the lower limit of the time window when passengers expect the airport bus to arrive, and the upper limit of the time window when passengers expect the airport bus to arrive;

[0049] The optimization model solving module is connected to the optimization model building module, and adopts a multi-layer coding optimization algorithm to solve the airport bus scheduling optimization model, outputting an airport bus operation route set, a departure schedule, and a passenger allocation plan to optimize the airport bus operation area network.

[0050] Optionally, the site location determination module is configured to:

[0051] Merge adjacent high-frequency demand grid cells to form a first passenger flow hotspot, and then perform a secondary merger on every two first passenger flow hotspots whose centroid distance is less than or equal to a preset distance threshold;

[0052] The first passenger flow hot zone after the second merger is divided and adjusted so that the horizontal span distance and the vertical span distance of the adjusted hot zone do not exceed the preset maximum service scale, thereby forming a second passenger flow hot zone;

[0053] Each high-frequency demand grid cell in the second passenger flow hot zone is scored according to the number of adjacent grids of each high-frequency demand grid cell in the second passenger flow hot zone, and the high-frequency demand grid cell center with the highest score result is selected as the airport bus stop location.

[0054] In summary, the technical solutions in the disclosed embodiments of the present invention can bring the following beneficial effects:

[0055] (1) Comprehensive consideration of multiple factors, including flight dynamics, actual passenger needs, and real-time urban traffic conditions, has enabled dynamic and flexible adjustments to airport bus routes and departure times. This adjustment not only improves the operational efficiency of airport buses, but also enhances their competitiveness in the urban transportation system, giving them a greater advantage in competition with other modes of transportation.

[0056] (2) By adopting the reservation response model, the operator can accurately obtain the travel needs of passengers before they travel. In particular, when there are changes in flights, such as delays, cancellations, and other special circumstances, the operator can quickly adjust the operation strategy based on this demand information. In this way, the time passengers wait for the bus and the time they spend on the bus can be effectively shortened, thereby significantly improving passenger satisfaction with the airport bus service;

[0057] (3) By accurately identifying and analyzing the locations of alternative airport bus stops, the layout of the airport bus network can be optimized to make it more scientific and reasonable. Such layout adjustments can better cover hot spots where passenger travel is concentrated, expand the coverage of airport bus services, and allow more passengers to conveniently enjoy the airport bus service.

[0058] Other features and advantages disclosed in the present invention will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The accompanying drawings are used to provide a further understanding of the present disclosure and constitute a part of the specification. Together with the following detailed description, they are used to explain the present disclosure but do not constitute a limitation of the present disclosure. In the accompanying drawings:

[0060] Figure 1This is a flow chart illustrating a method for optimizing an airport bus network by integrating air traffic control information according to an exemplary embodiment;

[0061] Figure 2 is based on Figure 1 A schematic flow chart of a method for determining a site location is shown;

[0062] Figure 3 It is a schematic diagram for identifying high-frequency demand grid cells;

[0063] Figure 4 It is a schematic diagram of hot zone merging;

[0064] Figure 5 It is a schematic diagram of the travel behavior characteristics of online ride-hailing passengers;

[0065] Figure 6 It is a schematic diagram of the spatial distribution of alternative airport bus stop locations;

[0066] Figure 7 It is a schematic diagram of the airport bus route distribution in a planning scheme;

[0067] Figure 8 The present invention is a structural diagram of an airport bus network optimization system integrating air traffic control information according to an exemplary embodiment. DETAILED DESCRIPTION

[0068] The following is a detailed description of the specific embodiments disclosed in the present invention in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are only used to illustrate and explain the present invention and are not intended to limit the present invention.

[0069] Figure 1 FIG. 1 is a flow chart showing a method for optimizing an airport bus network by integrating air traffic control information according to an exemplary embodiment. Figure 1 As shown, the method includes:

[0070] In step 101, the geographical scope of the airport bus operating area, the passenger reservation demand caused by the dynamic changes of flights, and the passenger online car-hailing order data are obtained based on the air traffic control information.

[0071] In the disclosed embodiment of the present invention, in view of the dynamic changes of flights and combined with the operational characteristics of airport buses, an airport bus network optimization method integrating air traffic control information is proposed, focusing on solving the following key problems: for flights with flight changes, a reservation response method is adopted to accurately obtain information such as the number of passengers taking the airport bus and their destinations; the order data of airport online car-hailing is used to identify alternative airport bus stops, providing a reasonable stop layout for the subsequent airport bus network design; the airport bus scheduling problem is converted into a semi-flexible demand response problem, which comprehensively considers flight dynamic information, passenger reservation data and urban traffic conditions, optimizes the operation routes and departure times of airport buses, and ensures the flexibility and efficiency of airport bus operations.

[0072] Among them, the passenger reservation requirements include: number of passengers, reservation time window and destination.

[0073] For example, the geographic scope of the bus operation area is obtained by loading a predefined geospatial data file. The obtained information contains the precise geographic boundaries of the geographic area, which is recorded as .

[0074] Passenger reservation demands generated by aviation dynamics include information such as the number of passengers, reservation time window, and destination. Each passenger reservation demand is recorded as a set , so the passenger reservation demand under the influence of aviation dynamics can be recorded as the set .

[0075] Collect the passenger online car-hailing order data and extract the geographical coordinates of the passenger drop-off location corresponding to each valid order record, which is recorded as , so the passenger online car-hailing order data can be recorded as a set It should be noted that the passenger online car-hailing order data collected in the disclosed embodiment of the present invention is electronic order data, including taxi orders and online car-hailing orders registered for private cars.

[0076] In step 102, the operation area is gridded to divide the operation area into an airport bus operation area network consisting of a plurality of grid units.

[0077] For example, with respect to step 102, if the airport bus stop location layout plan has been completed for the airport bus operating area before executing step 102, the planned airport bus stop locations can be directly obtained, and steps 102-104 can be skipped. After executing step 101, the process can be directly transferred to steps 105 and 106 to schedule the airport bus. If the airport bus stop location plan has not been completed within the geographic area, alternative airport bus stops can be identified and an airport bus scheduling optimization model can be constructed and solved according to the following steps 102-106.

[0078] For example, in order to facilitate subsequent research, the geographical area is gridded: a size parameter representing the spatial resolution of the grid cell is set, that is, the preset side length of the grid cell, which is recorded as Based on the precise geographic boundaries of the study area (i.e., the geographic area where the airport shuttle operates) and preset side lengths of the grid cells ,pass The function calculates and generates grid cells covering the study area, and sets the UTM coordinates of each grid cell , Datum Origin , precise size Record as a collection Therefore, all grids covering the study area are recorded as a set .

[0079] In step 103, the passenger flow demand of each grid unit is counted based on the distribution of the drop-off locations in the passenger's online car-hailing order data, and high-frequency demand grid units with passenger flow demand higher than a preset demand threshold are screened out.

[0080] For example, a grid index mapping operation is performed on the drop-off locations in the passenger's online car-hailing order data to determine the grid cells corresponding to each drop-off location. ; Count the total number of passenger drop-off events in the grid unit within the preset time period as the passenger flow demand of the grid unit ; Passenger flow demand Higher than the preset demand threshold Grid cells Determined as a high-frequency demand grid unit ; Get the high-frequency demand grid unit set as .

[0081] Specifically, define a deterministic mapping function from geographic coordinates to raster indexes , according to the origin, cell size and direction of the grid system, the geographic coordinates ( ) is converted into a set of index numbers that uniquely identify the grid cells to which the drop-off location belongs ( ), the index number is composed of the grid column number in the longitude direction and the grid row number in the latitude and longitude directions constitute.

[0082] Passenger online car-hailing order data collection By performing the grid index mapping operation, the rasterization of the drop-off location data can be completed, so that the drop-off location in each order is accurately assigned to a specific geographic grid cell. For each grid cell that successfully matches the passenger drop-off location (the grid cell is identified by its unique column number and row number), count the total number of passenger drop-off events carried by the grid cell in a preset time period, denoted as grid cell Passenger flow demand in a specific period Set a preset demand threshold Compare the passenger flow demand of all grid cells that have completed passenger flow counting with the preset threshold. If the passenger flow demand of a grid cell meets the condition , the grid cell is identified as a high-frequency demand grid cell, as shown in . The UTM coordinates , reference origin , and accurate size Figure 3 of the identified high-frequency demand grid cell are recorded as the high-frequency demand grid cell , so the high-frequency demand grid set can be denoted as . In step 104, the airport bus station location in the operation area is determined according to the high-frequency demand grid cell. In detail,

[0083] is a flowchart of a station location determination method, as shown in , step 104 includes:

[0084] Figure 2 In step 1041, adjacent high-frequency demand grid cells are merged to form first passenger flow hot zones, and then every two first passenger flow hot zones with a center distance less than or equal to a preset distance threshold are merged again. Figure 1 Figure 2 For example, based on the high-frequency demand grid cells obtained in the above steps, a clustering method based on grid adjacency relationship is used to identify and aggregate spatially adjacent high-frequency demand grid cells to form preliminary passenger flow hot zones, and each independently connected area is defined as a first passenger flow hot zone

[0085] . The coordinates of each first passenger flow hot zone can be calculated by the following formula:

[0086] ,

[0087] ,

[0088] ,

[0089] wherein and are the center coordinates of the first passenger flow hot zone ,​​​ and They constitute the first passenger flow hot zone High frequency demand grid unit The UTM coordinates, To form the first passenger flow hot zone High frequency demand grid unit passenger flow demand.

[0090] On the basis of forming the first passenger flow hot zone, set a preset distance threshold , for any two first passenger flow hot zones and ,according to Calculate the centroid distance between the first passenger flow hot zones , if the center of gravity distance between the two is less than or equal to the preset distance threshold , then these two first passenger flow hot zones are merged to form a new, larger first passenger flow hot zone, such as Figure 4 This process is repeated until all first passenger flow hotspots that meet the merging conditions are merged.

[0091] In step 1042, the first passenger flow hot zone after the second merger is divided and adjusted so that the horizontal span distance and the vertical span distance of the adjusted hot zone do not exceed the preset maximum service scale, thereby forming a second passenger flow hot zone.

[0092] Specifically, calculate the maximum horizontal span distance of each first passenger flow hot zone and maximum vertical span distance If the first passenger flow hot zone Maximum horizontal span distance Greater than , then the first passenger flow hot zone is horizontally Perform the segmentation, where is the preset maximum service scale; if the first passenger flow hot zone Maximum vertical span distance Greater than , then the first passenger flow hot zone is vertically Segmentation; if the first passenger flow hot zone The maximum vertical span distance and the maximum horizontal span distance are greater than , then the first passenger flow hot zone is located in the direction with the largest span distance. Split; determine the split position based on the principle of minimizing the absolute value of the sum of the passenger flow demand of all grid cells in the two hot zones formed after the split; and split the first passenger flow hot zone according to the split direction and split position. Split until the maximum horizontal span distance of all hot zones after splitting and maximum vertical span distance All less than , get the second passenger flow hot zone The second passenger flow hot zone set within the geographical area where the airport bus operates is recorded as .

[0093] In step 1043, each high-frequency demand grid cell in the second passenger flow hot zone is scored according to the number of adjacent grid cells of each high-frequency demand grid cell in the second passenger flow hot zone, and the high-frequency demand grid cell center with the highest score is selected as the airport bus stop location.

[0094] Specifically, the method for scoring the high-frequency demand grid cells in each second passenger flow hot zone is as follows: ,in, For the scoring results, The second passenger flow hot zone The number of adjacent grids of a high-frequency demand grid cell ( ), is the passenger flow demand of the second passenger flow hot zone, and is the weight coefficient, Represents each grid cell, Represents a high-frequency demand grid cell.

[0095] After obtaining the scoring results for each high-frequency demand grid cell, these high-frequency demand grid cells are then sorted from highest to lowest according to the numerical value of the scoring results. The high-frequency demand grid cell with the highest score is determined as the grid cell where the candidate airport bus stop location is located, and the coordinates of the center point of this grid cell are selected as the candidate airport bus stop location. The above operation is repeated one by one for all second passenger flow hotspots to accurately determine the layout location of all airport bus stops. At the same time, the passenger flow information between each second passenger flow hotspot is also accurately matched with the location of each airport bus stop.

[0096] In step 105, an airport bus scheduling optimization model is constructed.

[0097] Among them, the objective function of the airport bus scheduling optimization model is the sum of the minimized passenger travel time cost and the minimized bus operating cost. The constraints of the airport bus scheduling optimization model are established based on the maximum number of service passengers at the airport bus starting station, the service level of the airport bus schedule, the lower limit of the time window when passengers expect the airport bus to arrive, and the upper limit of the time window when passengers expect the airport bus to arrive.

[0098] For example, the costs that need to be considered in the design of the airport bus network include passenger travel time cost and airport bus operating cost. Therefore, an aviation dynamics-based reservation response bus scheduling model is established with the goal of minimizing the sum of passenger travel time cost and airport bus operating cost.

[0099] Specifically, the objective function is established based on the passenger travel time cost and the airport bus operating cost. ,in, is the objective function, The travel time cost of passengers, For the operating costs of the airport shuttle, and is the weight coefficient, and the calculation formula for the passenger's travel time cost is: ,in, , , , The time required for passengers to walk from the airport to the airport bus stop; The time the passenger is on board, To deviate from the time cost of service, Indicates passenger reservation demand With the airport bus If the passenger's reservation requirements By Airport Bus Providing services ,otherwise , is the distance between the passenger's airport and the airport bus stop, The average speed at which passengers walk to the airport bus stop. Indicates airport bus The starting point location and airport bus stop location The distance between Indicates airport bus From site location Drive to the station location The average running speed, Airport shuttle for passengers Arrival at the site The lower limit of the time window, The location of the airport bus stop that passengers expect to arrive at The upper limit of the time window, For airport buses Passengers arrive at the station location The calculation formula for the airport bus operating cost is: ,in, , , , , is the fixed cost of the vehicle, is the variable cost of the operation, is the penalty cost of the airport bus deviating from the passenger expected arrival time, is the penalty cost generated by the system rejecting an order, denotes the airport bus whether the airport bus goes from the origin site location to the site location if yes, then if no, is the depreciation coefficient of the airport bus for each operation, is the fuel cost per unit mileage of the airport bus vehicle, is the cost per unit mileage of the airport bus driver driving, is the penalty coefficient of the airport bus arriving at the site location earlier than the passenger expected arrival time, is the penalty coefficient of the airport bus arriving at the site location later than the passenger expected arrival time, is the penalty cost coefficient generated by the system rejecting an order, is the set of airport buses, is the set of airport bus site locations.

[0100] The first constraint condition is established according to the maximum number of passengers served by the airport bus origin site location: wherein, is the maximum number of passengers served by the airport bus origin site location, denotes the relationship between the passenger reservation demand and the airport bus , if the passenger reservation demand is served by the airport bus , then , otherwise , is the set of passenger reservation demands, is the set of airport buses;

[0101] The second constraint condition is established according to the service level of the airport bus schedule: wherein, is the minimum service level of the airport bus , is the maximum service level of the airport bus , denotes whether the airport bus goes from the origin site location to the site location , if yes, then , if no, , Gather for the location of the airport bus stop;

[0102] The third constraint is established based on the lower limit of the time window when passengers expect the airport bus to arrive and the upper limit of the time window when passengers expect the airport bus to arrive: ,in, The location of the airport bus stop that passengers expect to arrive at The lower limit of the time window, The location of the airport bus stop that passengers expect to arrive at The upper limit of the time window.

[0103] It is understandable that the airport bus schedule is restricted. The capacity is to ensure that the demand for each airport bus does not exceed the total carrying capacity of its actual vehicle; The service level of the airport bus is to ensure The service quality of each trip; constraining the airport bus to arrive at the passenger drop-off point within the upper limit of the time window when the passenger expects the airport bus to arrive.

[0104] In step 106, a multi-layer coding optimization algorithm is used to solve the airport bus scheduling optimization model, and the airport bus operation route set, departure schedule, and passenger allocation plan are output to optimize the airport bus operation area network.

[0105] For example, in order to achieve an efficient solution to the reservation-response bus scheduling model, the disclosed embodiment of the present invention proposes an optimization algorithm based on Adaptive Large Neighborhood Search (ALNS). By dynamically adjusting the combination of destruction and repair operations, the algorithm flexibly explores the solution space and quickly generates a low-cost, high-service quality scheduling solution in a scenario where flights are dynamically changing.

[0106] Specifically, a three-layer coding approach is used to design the solution: the airport bus set Defined as the first level encoding, where For airport buses The shift vehicle number; the service appointment request sequence is defined as the second level encoding, where For airport buses Service reservation request; Sequence the departure time Defined as the third level encoding, where For airport buses The vehicle number of the shift The corresponding departure time of the starting station; the initial feasible solution is designed and generated by using a three-layer coding method .

[0107] Passenger demand allocation: Use a greedy strategy to allocate each passenger's reservation demand Assigned to airport bus , giving priority to satisfying time window constraints and service capacity limitations. The allocation rules are:

[0108]

[0109] Based on the passenger demand allocation results, the Dijkstra algorithm is used to generate the initial airport bus route and calculate the departure time under the shortest path .

[0110] Feasibility verification: Check and generate initial feasible solutions Whether the constraints established in the above step 103 are met, if there is a conflict, it is repaired by adjusting the departure time or reallocating passengers until a feasible solution that meets the constraints is obtained.

[0111] Define multiple sets of destruction operations (Destroy) and repair operations (Repair), and select the optimal combination of operations through a dynamic weight adjustment mechanism:

[0112] Design four types of destruction operators To expand the neighborhood of candidate solutions, specifically including:

[0113] Random removal operator , indicating that from the current solution Randomly remove Passenger reservation needs , and get the solution after random removal And add the removed appointment requirements back to the list of pending requirements middle.

[0114] Maximum time window conflict removal operator : Define time window conflicts , first remove the passengers with the most serious time window conflict, and get the solution after removing the largest time window conflict And add the removed appointment requirements back to the list of pending requirements middle.

[0115] Inefficient path removal operator :in accordance with Filter out the flights with the highest path cost , treat it as an inefficient path and remove it, and get the solution after inefficient removal And add the removed appointment requirements back to the list of pending requirements middle.

[0116] Worst Damage Operator : Define the cost of damage ,in Indicates the current solution The total cost, Represents a hypothetical solution Remove appointment request The total cost after the destruction of After sorting in descending order, select the first The reservation request is removed and the solution after the worst damage is removed is obtained. And add the removed appointment requirements back to the list of pending requirements middle.

[0117] Solution after destruction and a list of requirements to be responded to , two repair operators are designed To speed up the neighborhood search process, specifically:

[0118] Random repair operator : Traverse the list of requirements to be responded to Each reservation request , insert it randomly into Update airport bus collection , Service Appointment Request Sequence and departure time sequence . Until the list of requirements to be responded to is completed All the reservation requests in the .

[0119] Greedy repair operator : Define the cost of repair ,in Indicates the current solution The total cost, Represents a hypothetical solution Insert a list of pending requests Appointment Requests The total cost after the repair After sorting in ascending order, the method that is ranked first is selected for repair according to the principle of minimum cost change, and the new solution after repair is obtained. , update airport bus collection , Service Appointment Request Sequence and departure time sequence . Until the list of requirements to be responded to is completed All the reservation requests in the .

[0120] The destruction operator designed in the above steps and repair operators Randomly combine two by two to generate Group operator pairs .

[0121] Define each operator pair The initial weight , the initial score is .

[0122] Repeat the above steps for iterative optimization until the stopping criterion of the algorithm is met.

[0123] According to the updated operator pair weights, the combination of destruction and repair operations for the next iteration is selected based on the roulette strategy.

[0124] According to the determined operator pair , we can see that this iteration selects the destruction operator and repair operators . According to the destruction and repair operations, the current solution Operate and get new solutions .

[0125] represent the current solution, new solution and optimal solution respectively, is the objective function value. ,but ;like ,but ;like , then the simulated annealing criterion is used to determine whether to accept the new solution. Calculate the simulated annealing probability , generate random numbers ,like ,but Otherwise, the new interpretation is not accepted. Specifically, the annealing probability The calculation formula is:

[0126] .

[0127] After each iteration, the score is updated according to the improvement degree of the operation on the objective function. The resulting solution after the destruction repair operation The new global optimal solution , Score increase ; If the operator The resulting solution after the destruction repair operation For better candidate solutions , Score increase ; If the operator The resulting solution after the destruction repair operation is a new feasible solution but does not improve the objective function value, Score increase .

[0128] Every In iterations, the weights of the operator pairs are updated. Specifically, the operator pair update formula is:

[0129]

[0130] in, is the learning rate.

[0131] The algorithm terminates when any of the following conditions are met: (1) the maximum number of iterations is reached; (2) the number of consecutive The global optimal solution is not improved by the iterations; (3) the computation time exceeds the threshold.

[0132] After the termination condition is triggered, the algorithm terminates and outputs the optimal scheduling solution . This includes the airport bus route set, departure schedule, passenger allocation plan and objective function value in the scheduling plan.

[0133] In addition, some defined variables in the embodiments disclosed in the present invention need to be explained:

[0134] Defining 0-1 variables If the airport bus From the starting site location Drive to the station location ,but , otherwise it is equal to The site location Belongs to the airport bus station location set N, airport bus Belong to the airport bus collection .

[0135] Defining 0-1 variables Indicates passenger reservation demand With the airport bus If the passenger reservation demand By Airport Bus Provide services, then ,otherwise .

[0136] Defining continuous variables Indicates airport bus Arrival at the site moment.

[0137] Defining continuous variables Indicates airport bus From the starting site location Drive to the station location Average running speed, m / s.

[0138] Example:

[0139] The data used in this example are 1,726,511 online ride-hailing order records in a certain city, collected from some online ride-hailing platforms within a specific time period. This data is based on actual operation records, has high temporal and spatial resolution and good data integrity, and can more realistically reflect the travel behavior characteristics of online ride-hailing passengers in a certain city, such as Figure 5 As shown. The order data with the airport as the departure point is filtered as the passenger online car-hailing order data to identify the location of alternative airport bus stops and provide a reasonable stop layout for the subsequent airport bus network design. Finally, 81 alternative express bus stops are screened out. Figure 6 The spatial distribution of the alternative airport bus stops. The booking platform has received 18 passenger booking requests influenced by flight dynamics.

[0140] like Figure 7 As shown, the planned airport bus route is 38.65 km long, with an estimated travel time of 45 minutes. From a passenger perspective, this travel time is acceptable for medium- and long-distance trips. Comparing the route's linear distance with the airport bus's operating distance reveals a total detour distance of 12.53 km, with an average detour distance of 1.14 km. The route's non-linearity coefficient is 1.43, indicating that the route covers a wide area and the degree of detour is acceptable. For example, a comparison of the arrival times of passengers taking express buses and ride-hailing services at an average speed of 35 km / h reveals that the maximum delay for passengers taking the airport bus is 18.76 minutes, accounting for approximately one-quarter of the passenger's total travel time, with an average delay of 1.52 minutes. This demonstrates the overall operational efficiency of the route, with minimal additional waiting time for passengers. The optimized route features 14 stops, meeting the maximum number of stops allowed. The average distance between stops is 2.78 km. The optimal route can meet the needs of 88.89% of passengers, indicating that the service quality of the route is high and can effectively improve the passengers' travel experience.

[0141] The above results also show that the airport bus network optimization method integrating air traffic control information proposed in the present invention has good characteristics and can be extended to practical applications.

[0142] Figure 8 FIG. 1 is a schematic diagram showing a structure of an airport bus network optimization system integrating air traffic control information according to an exemplary embodiment. Figure 8 As shown, the system 800 includes:

[0143] Data acquisition module 810 acquires the geographical scope of the airport shuttle operation area, passenger reservation demands caused by flight dynamic changes, and passenger online ride-hailing order data based on air traffic control information. The passenger reservation demands include: number of passengers, reservation time window, and destination;

[0144] A rasterization processing module 820 is connected to the data acquisition module 810 and performs rasterization processing on the operation area to divide the operation area into an airport bus operation area network consisting of a plurality of raster units;

[0145] A high-frequency grid determination module 830 is connected to the grid processing module 820 and calculates the passenger flow demand of each grid cell based on the distribution of the drop-off locations in the passenger online ride-hailing order data, and selects high-frequency demand grid cells whose passenger flow demand exceeds a preset demand threshold;

[0146] A station location determination module 840 is connected to the high-frequency grid determination module 830 and determines the location of the airport bus station within the operating area based on the high-frequency demand grid unit;

[0147] An optimization model construction module 850 is connected to the station location determination module 840 and constructs an airport bus scheduling optimization model. The objective function of the airport bus scheduling optimization model is the sum of the minimized passenger travel time cost and the minimized bus operating cost. The constraints of the airport bus scheduling optimization model are established based on the maximum number of service passengers at the airport bus starting station, the service level of the airport bus schedule, the lower limit of the time window when passengers expect the airport bus to arrive, and the upper limit of the time window when passengers expect the airport bus to arrive.

[0148] The optimization model solving module 860 is connected to the optimization model building module 850, and adopts a multi-layer coding optimization algorithm to solve the airport bus scheduling optimization model, outputting the airport bus operation route set, departure schedule, and passenger allocation plan to optimize the airport bus operation area network.

[0149] Optionally, the site location determination module 840 is configured to:

[0150] Merge adjacent high-frequency demand grid cells to form a first passenger flow hotspot, and then perform a secondary merger on every two first passenger flow hotspots whose centroid distance is less than or equal to a preset distance threshold;

[0151] The first passenger flow hot zone after the second merger is divided and adjusted so that the horizontal span distance and the vertical span distance of the adjusted hot zone do not exceed the preset maximum service scale, thereby forming a second passenger flow hot zone;

[0152] Each high-frequency demand grid cell in the second passenger flow hot zone is scored according to the number of adjacent grids of each high-frequency demand grid cell in the second passenger flow hot zone, and the high-frequency demand grid cell center with the highest score result is selected as the airport bus stop location.

[0153] In summary, the present invention discloses an airport bus network optimization method and system that integrates air traffic control information. The method includes: obtaining the geographical scope of the airport bus operating area, passenger reservation demand and passenger online car-hailing order data generated by flight dynamic changes based on air traffic control information; determining the locations of all airport bus stops within the airport bus operating area; establishing an objective function based on passenger travel time cost and airport bus operating cost, and establishing constraints based on the maximum number of service passengers at the airport bus starting station, the service level of the airport bus schedule, the lower limit of the time window when passengers expect the airport bus to arrive, and the upper limit of the time window when passengers expect the airport bus to arrive; solving the reservation response bus scheduling model to obtain an airport bus scheduling plan. In the event of flight changes, the method can quickly adjust the operating strategy and optimize vehicle scheduling by knowing in advance the number of passengers taking the airport bus and their destinations, thereby effectively saving passengers' waiting time and time on the bus, and thus improving passenger satisfaction in flight conditions.

[0154] The preferred embodiments of the present disclosure are described in detail above in conjunction with the accompanying drawings. However, the present disclosure is not limited to the specific details of the above embodiments. Within the technical concept of the present disclosure, various simple modifications can be made to the technical solutions of the present disclosure, and these simple modifications all fall within the scope of protection of the present disclosure.

[0155] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any appropriate manner without contradiction. In order to avoid unnecessary repetition, the present disclosure will not further describe various possible combinations.

[0156] In addition, the various embodiments of the present disclosure may be arbitrarily combined, and as long as they do not violate the concept of the present disclosure, they should also be regarded as the contents disclosed by the present disclosure.

Claims

1. An airport bus network optimization method integrating air traffic control information, characterized in that: The method comprises: Obtain the geographical scope of the airport shuttle operation area, passenger reservation demands caused by flight dynamics, and passenger online ride-hailing order data based on air traffic control information. The passenger reservation demands include: number of passengers, reservation time window, and destination; Performing grid processing on the operation area to divide the operation area into an airport bus operation area network consisting of a plurality of grid units; Counting the passenger flow demand of each grid cell based on the distribution of the drop-off locations in the passenger online ride-hailing order data, and screening out high-frequency demand grid cells whose passenger flow demand is higher than a preset demand threshold; Determining the location of the airport bus stop within the operating area according to the high-frequency demand grid unit; Constructing an airport bus scheduling optimization model, wherein the objective function of the airport bus scheduling optimization model is the sum of minimized passenger travel time cost and minimized bus operating cost, and the constraints of the airport bus scheduling optimization model are established based on the maximum number of service passengers at the airport bus starting station, the service level of the airport bus schedule, the lower limit of the time window when passengers expect the airport bus to arrive, and the upper limit of the time window when passengers expect the airport bus to arrive; A multi-layer coding optimization algorithm is used to solve the airport bus scheduling optimization model, and an airport bus operation route set, a departure schedule, and a passenger allocation plan are output to optimize the airport bus operation area network; The determining of the location of the airport bus stop within the operating area according to the high-frequency demand grid unit includes: Merge adjacent high-frequency demand grid cells to form a first passenger flow hotspot, and then perform a secondary merger on every two first passenger flow hotspots whose centroid distance is less than or equal to a preset distance threshold; The first passenger flow hot zone after the second merger is divided and adjusted so that the horizontal span distance and the vertical span distance of the adjusted hot zone do not exceed the preset maximum service scale, thereby forming a second passenger flow hot zone; Scoring each high-frequency demand grid cell in the second passenger flow hot zone according to the number of adjacent grid cells of each high-frequency demand grid cell in the second passenger flow hot zone, and selecting the high-frequency demand grid cell center with the highest score as the location of the airport bus stop; The multi-layer coding optimization algorithm is used to solve the airport bus scheduling optimization model, including: Gather the airport bus Defined as the first level encoding, where For airport buses The vehicle number of the shift; Sequence the service appointment request is defined as the second level encoding, where For airport buses service appointment requests; Departure time sequence Defined as the third level encoding, where For airport buses The vehicle number of the shift The corresponding departure time from the starting station; The initial feasible solution is designed and generated using a three-layer coding method. .

2. The airport bus network optimization method integrating air traffic control information according to claim 1 is characterized in that: The method of counting passenger flow demand of each grid unit based on the distribution of drop-off locations in the passenger online ride-hailing order data, and screening out high-frequency demand grid units whose passenger flow demand is higher than a preset demand threshold, includes: Perform a grid index mapping operation on the drop-off locations in the passenger's online ride-hailing order data to determine the grid cells corresponding to each drop-off location ; Count the total number of passenger drop-off events in the grid unit within a preset time period as the passenger flow demand of the grid unit ; Passenger flow demand Higher than the preset demand threshold Grid cells Determined as a high-frequency demand grid unit ; Get the high-frequency demand grid cell set as .

3. The airport bus network optimization method integrating air traffic control information according to claim 1 is characterized in that: The dividing and adjusting the first passenger flow hot zone after the second merger so that the horizontal span distance and the vertical span distance of the adjusted hot zone do not exceed the preset maximum service scale to form a second passenger flow hot zone includes: Calculate the maximum horizontal span distance of each first passenger flow hot zone and maximum vertical span distance ; If the first passenger flow hot zone Maximum horizontal span distance Greater than , then the first passenger flow hot zone is horizontally Perform the segmentation, where The preset maximum service scale; If the first passenger flow hot zone Maximum vertical span distance Greater than , then the first passenger flow hot zone is vertically Perform segmentation; If the first passenger flow hot zone The maximum vertical span distance and the maximum horizontal span distance are greater than , then the first passenger flow hot zone is Perform segmentation; The splitting position is determined based on the principle of minimizing the absolute value of the difference between the sum of passenger flow demands of all grid cells in the two hot zones formed after the splitting. The first passenger flow hot zone is divided according to the segmentation direction and segmentation position. Split until the maximum horizontal span distance of all hot zones after splitting and maximum vertical span distance All less than , get the second passenger flow hot zone ; The second passenger flow hot zone set within the geographical area where the airport bus operates is recorded as .

4. The airport bus network optimization method integrating air traffic control information according to claim 1 is characterized in that: The objective function of the airport bus scheduling optimization model is expressed as: ,in, is the objective function, The travel time cost of passengers, For the operating costs of the airport shuttle, and is the weight coefficient, and the calculation formula for the passenger travel time cost is: ,in, , , , The time required for passengers to walk from the airport to the airport bus stop; The time the passenger is on board, To deviate from the time cost of service, Indicates passenger reservation demand With the airport bus If the passenger's reservation requirements By Airport Bus Providing services ,otherwise , is the distance between the passenger's airport and the airport bus stop, represents the average speed at which passengers walk to the airport bus stop. Indicates airport bus The starting point location and airport bus stop location The distance between Indicates airport bus From site location Drive to the station location The average running speed, Airport shuttle for passengers Arrival at the site The lower limit of the time window, The location of the airport bus stop that passengers expect to arrive at The upper limit of the time window, For airport buses Passengers arrive at the station location The calculation formula of the airport bus operating cost is: ,in, , , , , is the fixed cost of the vehicle, For variable operating costs, The penalty cost for the airport bus to deviate from the passenger's expected arrival time is: The penalty cost for the system rejecting the order, Indicates airport bus Whether to go from the starting station location to the station location If so, , if otherwise , is the depreciation factor for each operation of the airport bus, is the fuel cost per mile for the airport shuttle vehicle, The unit mileage cost for airport shuttle drivers, is the penalty coefficient for the actual arrival time of the airport bus being earlier than the passenger’s expected arrival time at the station, is the penalty coefficient for the actual arrival time of the airport bus being later than the passenger’s expected arrival time at the station. The penalty cost coefficient generated by the system rejecting the order, Gathering for the airport bus, Gather at the airport bus stop location.

5. The airport bus network optimization method integrating air traffic control information according to claim 1 is characterized in that: The process of establishing the constraint conditions includes: The first constraint is established based on the maximum number of service passengers at the starting stop of the airport bus: ,in, The maximum number of passengers served at the starting point of the airport bus. Indicates passenger reservation demand With the airport bus If the passenger's reservation requirements By Airport Bus Providing services ,otherwise , Collect reservation requirements for passengers, Gather for the airport bus; The second constraint is established based on the service level of the airport bus service: ,in, For airport buses Minimum service level, For airport buses Maximum service level, Indicates airport bus Whether to go from the starting station location to the station location If so, , if otherwise , Gather for the location of the airport bus stop; The third constraint is established based on the lower limit of the time window when passengers expect the airport bus to arrive and the upper limit of the time window when passengers expect the airport bus to arrive: ,in, The location of the airport bus stop that passengers expect to arrive at The lower limit of the time window, The location of the airport bus stop that passengers expect to arrive at The upper limit of the time window.

6. An airport bus network optimization system integrating air traffic control information, characterized in that: The system comprises: A data acquisition module, which obtains the geographical scope of the airport shuttle operation area, passenger reservation demands caused by flight dynamics, and passenger online ride-hailing order data based on air traffic control information. The passenger reservation demands include the number of passengers, reservation time window, and destination. a rasterization processing module connected to the data acquisition module, performing rasterization processing on the operation area so as to divide the operation area into an airport bus operation area network consisting of a plurality of raster units; a high-frequency grid determination module, connected to the grid processing module, which calculates the passenger flow demand of each grid cell based on the distribution of the drop-off locations in the passenger online ride-hailing order data, and selects high-frequency demand grid cells whose passenger flow demand exceeds a preset demand threshold; a station location determination module, connected to the high-frequency grid determination module, and determining the location of the airport bus station within the operating area according to the high-frequency demand grid unit; an optimization model construction module, connected to the station location determination module, for constructing an airport bus scheduling optimization model, wherein the objective function of the airport bus scheduling optimization model is the sum of minimized passenger travel time cost and minimized bus operating cost, and the constraints of the airport bus scheduling optimization model are established based on the maximum number of service passengers at the airport bus starting station, the service level of the airport bus schedule, the lower limit of the time window when passengers expect the airport bus to arrive, and the upper limit of the time window when passengers expect the airport bus to arrive; an optimization model solving module, connected to the optimization model building module, using a multi-layer coding optimization algorithm to solve the airport bus scheduling optimization model, outputting an airport bus operation route set, a departure schedule, and a passenger allocation plan, so as to optimize the airport bus operation area network; The site location determination module is further configured to: Merge adjacent high-frequency demand grid cells to form a first passenger flow hotspot, and then perform a secondary merger on every two first passenger flow hotspots whose centroid distance is less than or equal to a preset distance threshold; The first passenger flow hot zone after the second merger is divided and adjusted so that the horizontal span distance and the vertical span distance of the adjusted hot zone do not exceed the preset maximum service scale, thereby forming a second passenger flow hot zone; Scoring each high-frequency demand grid cell in the second passenger flow hot zone according to the number of adjacent grid cells of each high-frequency demand grid cell in the second passenger flow hot zone, and selecting the high-frequency demand grid cell center with the highest score as the location of the airport bus stop; The optimization model solving module is also used for: Gather the airport bus Defined as the first level encoding, where For airport buses The vehicle number of the shift; Sequence the service appointment request is defined as the second level encoding, where For airport buses service appointment requests; Departure time sequence Defined as the third level encoding, where For airport buses The vehicle number of the shift The corresponding departure time from the starting station; The initial feasible solution is designed and generated using a three-layer coding method. .

Images