An airport passenger evacuation system based on modular autonomous bus shuttles

Abstract

The present invention discloses an airport passenger evacuation system based on modular autonomous driving bus connection, comprising: a data acquisition module, configured to fuse multi-source data to predict the number of people arriving at the remote parking lot side station in each time slot, and determine modular bus performance parameters, passenger time cost parameters and operation parameters; a scheduling plan generation module, based on the constructed scheduling model, inputting the parameters obtained by the data acquisition module, and solving the modular bus departure plan; an operation module, configured to automatically group and depart according to the modular bus departure plan, and complete the passenger connection between the remote parking lot and the airport terminal building. The present invention combines the concept of remote parking with the emerging modular autonomous driving bus technology, provides new ideas for the airport landside collection and distribution system, alleviates the problems of road congestion, difficult user parking and high costs around the airport terminal building, and at the same time improves the revenue of the remote parking lot operation, forms economies of scale to reduce energy consumption.

Description

Technical Field

[0001] The present invention relates to the technical field of modular autonomous driving bus operation, and in particular to an airport passenger evacuation system based on modular autonomous driving bus connection. Background Art

[0002] With the economic growth and technological development, air travel has become increasingly popular. With its characteristics of convenience, comfort and speed, it has become the choice for more and more people's long-distance travel. At the same time, the national car ownership has been rising continuously, and a considerable part of airport passengers choose to travel by private car, which brings great pressure to airport parking. Building a landside collection and distribution system that matches the airside facilities and functions of the airport is very important for enhancing the competitiveness of air travel. At present, there are the following pain points in the landside car travel at the airport: the land around the terminal is in short supply, and the construction of parking lots requires a large amount of land space. Therefore, there are few parking spaces near the terminal and the parking prices are expensive; during peak flight hours, the roads near the airport are prone to congestion, which may cause delays or even missed flights for passengers.

[0003] To relieve the traffic congestion and land shortage on the landside of the airport road and provide more preferential services for car travelers, remote parking has emerged, and thus the need for connection between the terminal and the remote parking lot has arisen. At present, the public remote parking lot uses shuttle buses to travel back and forth between the remote parking lot and the terminal. Due to different passenger peaks and troughs, this connection method causes some passengers to wait for a long time or results in waste of vehicle resources. The private operation mode is difficult to overall supervise, with low reliability, and the one-to-one service mode is also relatively inefficient.

[0004] Modular autonomous driving bus (referred to as modular bus for short) is a new hot spot in the transportation field. It is powered by electricity and based on autonomous driving technology. 6-8 seat small electric driverless buses can be grouped and dispatched according to the number of passengers, realizing flexible splicing and splitting, so as to improve the intelligence and accuracy of urban transportation services, and having great advantages in energy conservation, targeted services, etc. At present, the research on modular buses mainly focuses on the operation methods and algorithms of variable-capacity modular vehicles, and there is still a gap in the research under the application scenario of airport passenger evacuation systems.

[0005] Based on this, the present invention combines the concept of remote parking with the emerging modular autonomous driving bus technology, and innovatively proposes an airport passenger evacuation system based on modular autonomous driving bus connection, providing new ideas for the airport landside collection and distribution system. Compared with other traditional connection schemes for airport remote parking, modular buses are superior in terms of economy, environmental protection, low latency, etc., and conform to the construction concept of intelligent transportation. Summary of the Invention

[0006] The object of the present invention is to alleviate the problems of road congestion, shortage of parking spaces, difficult parking and high costs for users around the airport terminal. A passenger evacuation system for airports based on modular autonomous driving bus connection is provided to help the airport relieve the pressure on landside facilities, while increasing the revenue of the operation of the remote parking lot, achieving economies of scale and reducing energy consumption. By obtaining flight information and user reservation information and using a scheduling model to solve the optimal departure plan, modular autonomous driving buses can be automatically grouped and dispatched, thus completing the evacuation between the remote parking lot and the airport terminal and providing a more reliable and convenient connection service for passengers.

[0007] To achieve the above object, the technical solution provided by the present invention is: a passenger evacuation system for airports based on modular autonomous driving bus connection, including:

[0008] A data acquisition module, which is used to fuse multi-source data to predict the number of people arriving at the remote parking lot side station in each time slot, and determine the performance parameters of modular autonomous driving buses, the passenger time cost parameters and the operation parameters;

[0009] A scheduling plan generation module, based on the constructed scheduling model, inputs the parameters obtained by the data acquisition module, and solves the departure plan of modular autonomous driving buses, that is, when to depart and the number of vehicles in the departure formation, with the goal of minimizing the total economic cost of passengers and the operating agency;

[0010] An operation module, which is used to automatically group and dispatch according to the departure plan of the modular autonomous driving bus, and complete the passenger connection between the remote parking lot and the airport terminal.

[0011] Furthermore, in the data acquisition module, the multi-source data includes mini-program reservation data and airport flight data. It is stipulated that B = [0, 1, 2,..., J] is a set of discrete time points, J is the total number of time slots, and j is the discrete time point index. When j takes J, it represents the last discrete time point. The formula for calculating the number of people arriving at the remote parking lot side station in each time slot is:

[0012] a j = α × a 1j + β × a 2j

[0013] In the formula, a j represents the estimated number of passengers arriving at the station in the time slot [j - 1, j]; a 1j represents the estimated number of passengers arriving at the station in the time slot [j - 1, j] obtained through mini-program reservation data; a 2j represents the estimated number of passengers arriving at the station in the time slot [j - 1, j] obtained through airport flight data; α and β represent weights, and α + β = 1, α > 0, β > 0;

[0014] The estimated number of passengers arriving at the station at each time slot obtained through the mini-program reservation data can be obtained based on the arrival time at the parking lot filled in by the user on the remote parking lot reservation mini-program;

[0015] The estimated number of passengers arriving at the station at each time slot obtained from airport flight data is calculated as follows:

[0016] a 2j =g(S j *σ)

[0017] In the formula, S j represents the number of passengers arriving at the terminal in the future after the [j-1,j] time slot, which can be obtained based on the airport flight schedule, the passenger capacity of each flight and the scheduled occupancy rate; σ represents the passenger allocation ratio of modular autonomous driving buses to the terminal, which can be obtained based on historical operation data and survey data; function g fits the distribution of passengers arriving at the remote parking lot side station over time, and projects the number of people arriving at the terminal by modular autonomous driving buses in the future to the [j-1,j] time slot to obtain the number of passengers arriving at the remote parking lot side station in the [j-1,j] time slot. The calculation can adopt a second-order Gaussian mixture model, or it can be obtained based on survey data;

[0018] The weights can be obtained based on historical data by adopting a variety of optimization methods with the goal of minimizing the error between the calculated value and the actual value, and must be updated periodically.

[0019] Furthermore, in the data acquisition module, the modular autonomous driving bus performance parameters include the passenger capacity of a single modular vehicle, the fixed energy cost of a single departure, the energy cost coefficient of the number of vehicles in a single formation, and the energy calculation coefficient; the passenger time cost parameters include the waiting cost per passenger per unit time; and the operating parameters include the minimum departure time interval and the maximum number of vehicle formations.

[0020] Furthermore, in the scheduling scheme generation module, the scheduling model is applied to a passenger transfer system based on modular autonomous driving buses with a single starting and ending point to solve the departure plan of the modular autonomous driving buses at the starting point, that is, when to depart and the number of departure vehicles, with the goal of minimizing the total economic cost of passengers and operating agencies, including the following assumptions:

[0021] It is stipulated that A = [0, 1, 2, ..., I] is the vehicle composition set, I is the maximum number of vehicle groups, i is the vehicle composition index,

[0022] Adopt the discretization idea, divide the time range [0, T] of the departure plan to be solved into J integer time slots δ, where δ = T / J; it is stipulated that B = [0, 1, 2,..., J] is the set of discrete time points, J is the total number of time slots, and j is the discrete time point index. When j takes J, it represents the last discrete time point within the said time range.

[0023] The decision variable x ij is a binary variable, indicating whether to dispatch a modular autonomous driving bus with formation number i at time j, that is, to determine whether to depart and the number of vehicles in the corresponding departure formation at each discrete time point j.

[0024] Furthermore, the scheduling model includes an objective function, safety constraints, passenger number conservation constraints, and operation quality constraints.

[0025] The calculation formula of the objective function is:

[0026]

[0027] In the formula, cost total represents the total economic cost of passengers and the operation organization; cost agency represents the operation cost of the modular autonomous driving bus; cost passenger represents the waiting time cost of passengers; f i represents the total vehicle energy cost of a vehicle formation with formation number i for a single departure; x ij is a binary variable, indicating whether to dispatch a modular autonomous driving bus with formation number i at time j. If dispatched, then x ij = 1, otherwise x ij = 0; δ represents the fixed time slot length; w represents the average waiting cost per passenger per unit time; q j-1 represents the number of passengers waiting to board at time j - 1; a j represents the number of people arriving at the station on the remote parking lot side within the time slot [j - 1, j]; a 0 represents the initial number of waiting passengers; C F is the fixed energy cost for a single departure, C V is the energy cost coefficient for the number of vehicles in a single formation, μ ≤ 1 is the energy calculation coefficient, c i represents the total passenger capacity of a formation with vehicle number i, which is obtained by multiplying the number of vehicles in the formation i by the passenger capacity of a single modular autonomous driving bus.

[0028] The safety constraints are used to ensure that at most one formation of modular buses departs within one time slot. The formula is:

[0029]

[0030] The number conservation constraint is used to ensure the rationality of the calculation of the scheduling plan. Except for the number of people waiting at the departure time q 0 and the number of people waiting at the end time of the last bus q J , the number of people waiting at the end of the remaining time slots is the number of people waiting at the end of the previous time slot, plus the number of people newly arriving at the station on the remote parking lot side, minus the maximum value of the passenger capacity of the modular autonomous driving bus departing and 0. The formula is;

[0031] q 0 =a 0

[0032]

[0033] q J =0

[0034] The operation quality constraint includes the overcrowding constraint and the undercrowding constraint;

[0035] The overcrowding constraint is used to ensure that the number of people waiting in each time slot is not greater than the maximum number of people waiting Q that can be carried, so as to limit the longest waiting time of passengers. The formula is:

[0036]

[0037] The undercrowding constraint is used to ensure the departure frequency. A bus must depart every three time slots. The formula is:

[0038]

[0039] In the formula, x ij-1 , x ij-2 are both binary variables, indicating whether a modular autonomous driving bus with formation number i is dispatched at time j-1 and j-2. If dispatched, it takes 1, otherwise it takes 0.

[0040] Furthermore, the model of the scheduling plan generation module is solved by direct solution method, linear solution method or heuristic algorithm.

[0041] Furthermore, the operation module includes a reservation mini-program, a station on the remote parking lot side, a station on the airport terminal side, a modular autonomous driving bus, a modular autonomous driving bus lane between the remote parking lot and the terminal;

[0042] The reservation mini-program is used to collect the name of the remote parking lot, flight number, boarding gate, number of passengers, arrival date and specific time, and pay the parking fee online;

[0043] Both the remote parking lot side station and the airport terminal side station include the formation lines and storage lines of modular autonomous buses, passenger waiting areas, passenger dropping-off areas, and bus boarding toll collection points; the remote parking lot side station and the airport terminal side station also include corresponding parking services and terminal passenger services.

[0044] The modular autonomous bus lane between the remote parking lot and the terminal is set as a dedicated lane.

[0045] The modular autonomous buses are grouped and dispatched according to the generated modular autonomous bus scheduling plan, shuttling between the remote parking lot side station and the airport terminal side station; for multiple remote parking lots and multiple terminals, the modular autonomous buses depart from different remote parking lots and merge at the route intersection; when going to different terminals, they split at the route bifurcation to complete the passenger connection between the remote parking lot and the airport terminal.

[0046] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0047] 1. Introduce modular autonomous buses (abbreviated as modular buses) into the airport passenger evacuation system to provide a new solution for the connection between the remote parking lot and the airport terminal. Compared with the traditional connection method, the modular bus connection meets the requirements of intelligent transportation and has obvious advantages: the passenger travel experience is improved, the parking fee is reduced, and the waiting time is shortened; the airport passenger evacuation system can operate automatically, and the modular buses are automatically grouped, dispatched, and driven according to the passenger flow, reducing labor costs and management costs; the modular buses are grouped and dispatched according to the number of passengers, improving energy utilization efficiency, reducing carbon emissions and harmful gas emissions, and having good environmental benefits.

[0048] 2. Integrate multi-source data to predict the number of people arriving at the remote parking lot side station in each time slot, obtain airport flight data, consider the direct impact of flight schedules and the number of inbound and outbound passengers on the scheduling model, effectively alleviate the impact of the rapid change of flight schedules due to factors such as weather and traffic control, and combine the appointment data of the small program to improve the prediction accuracy of the number of people arriving at the remote parking lot side station in each time slot.

[0049] 3. Build a scheduling model and calculate the modular bus departure plan in real time. The scheduling model aims to minimize the total economic cost of passengers and operating agencies and includes three types of constraint conditions. By inputting the arrival number and other parameters, an efficient short-term formation scheduling plan can be generated to guide the splicing and unhooking of modular buses. Brief Description of the Drawings

[0050] Figure 1 It is the architecture diagram of the system of the present invention.

[0051] Figure 2 It is the schematic diagram of the appointment small program interface.

[0052] Figure 3 It is a schematic diagram of splicing and splitting during the connection of modular autonomous driving buses (hereinafter referred to as modular buses).

[0053] Figure 4 It is a system operation flow chart from the perspective of passengers. Specific implementation manners

[0054] The present invention will be further described in detail below in conjunction with embodiments and the accompanying drawings, but the implementation manners of the present invention are not limited thereto.

[0055] As Figure 1 shown, this embodiment discloses an airport passenger evacuation system based on the connection of modular autonomous driving buses, including the following modules:

[0056] A data acquisition module, which is used to fuse multi-source data to predict the number of people arriving at the remote parking lot side station in each time slot, and determine the performance parameters of modular autonomous driving buses (hereinafter referred to as modular buses), passenger time cost parameters and operation parameters;

[0057] A scheduling plan generation module, based on the constructed scheduling model, inputs the parameters obtained by the data acquisition module, and solves the departure plan of modular autonomous driving buses, that is, when to depart and the number of vehicles in the departure formation, with the goal of minimizing the total economic cost of passengers and operating agencies;

[0058] An operation module, which is used to automatically form a formation and depart according to the departure plan of the modular autonomous driving bus, and complete the passenger connection between the remote parking lot and the airport terminal building.

[0059] In the data acquisition module, the multi-source data includes mini-program reservation data and airport flight data. It is stipulated that B = [0, 1, 2,..., J] is a set of discrete time points, j is the discrete time point index, when j takes J, it represents the last discrete time point; the planned departure time of civil aviation flights in China is in units of five minutes, so the time slot is set to 5 minutes here. The calculation formula for the number of people arriving at the remote parking lot side station in each time slot is:

[0060] a j = α × a 1j + β × a 2j

[0061] where a j represents the estimated number of passengers arriving at the station in the time slot [j - 1, j]; a 1j represents the estimated number of passengers arriving at the station in the time slot [j - 1, j] obtained through mini-program reservation data; a 2jrepresents the estimated number of passengers arriving at the station in the [j-1, j] time slot obtained from the airport flight data; α and β represent weights, and α+β=1(α<0,β>0);

[0062] The estimated number of passengers arriving at the station at each time slot obtained through the mini-program reservation data can be obtained based on the arrival time at the parking lot filled in by the user on the remote parking lot reservation mini-program. The mini-program interface is shown as follows: Figure 2 shown.

[0063] The estimated number of passengers arriving at the station at each time slot obtained from airport flight data is calculated as follows:

[0064] a 2j =g(S j *σ)

[0065] Among them, S j represents the number of passengers arriving at the terminal in the future after the [j-1,j] time slot, which can be obtained based on the airport flight schedule, the passenger capacity of each flight and the scheduled occupancy rate; σ represents the passenger allocation ratio of modular buses to the terminal, which can be obtained based on historical operation data and survey data; function g fits the distribution of passengers arriving at the remote parking lot side station over time, and projects the number of people arriving at the terminal by modular buses in the future to the [j-1,j] time slot to obtain the number of passengers arriving at the remote parking lot side station in the [j-1,j] time slot. The calculation can adopt a second-order Gaussian mixture model or be obtained based on survey data;

[0066] Since there is no real mini-program reservation data, α and β are set to 0 and 1 respectively, and the estimated number of passengers is obtained only through airport flight data. The flight information of Guangzhou Baiyun Airport for one hour on a certain day is obtained from the "Air Travel" APP. σ is set to 10%, and the number of passengers expected to arrive at the station every 5 minutes is shown in Table 2.

[0067] The parameters of the data acquisition module also include modular bus performance parameters, passenger time cost parameters and operation parameters, as shown in Table 1; the modular bus performance parameters include the passenger capacity of a single modular vehicle, the fixed energy cost of a single departure, the energy cost coefficient of the number of vehicles in a single formation, and the energy calculation coefficient; the passenger time cost parameters include the waiting cost per passenger per unit time; the operation parameters include the minimum departure time interval and the maximum number of vehicle formations.

[0068] Table 1: Fixed parameter values

[0069]

[0070] In the scheduling plan generation module, the scheduling model is applied to a passenger transfer system based on modular autonomous driving buses with a single origin and destination to solve the departure plan of modular autonomous driving buses at the origin, that is, when to depart and the number of vehicles in the departure formation. Its goal is to minimize the total economic costs of passengers and operating agencies, including the following assumptions:

[0071] It is stipulated that \(A = [0, 1, 2, \cdots, I]\) is the set of vehicle formations, \(I\) is the maximum number of vehicle formations, and \(i\) is the vehicle formation index.

[0072] Adopting the discretization idea, the time range \([0, T]\) of the departure plan to be solved is divided into \(J\) integer time slots \(\delta\), where \(\delta=T / J\); it is stipulated that \(B = [0, 1, 2, \cdots, J]\) is the set of discrete time points, \(J\) is the total number of time slots, and \(j\) is the discrete time point index. When \(j\) takes \(J\), it represents the last discrete time point within the time range.

[0073] The decision variable \(x\) ij is a binary variable, indicating whether to dispatch a modular autonomous driving bus with a formation number of \(i\) at time \(j\), that is, to determine whether to depart at each discrete time point \(j\) and the number of vehicles in the corresponding departure formation.

[0074] The scheduling model includes an objective function, safety constraints, passenger number conservation constraints, and operation quality constraints.

[0075] The calculation formula of the objective function is as follows:

[0076]

[0077] In the formula, \(cost\) total represents the total economic cost of passengers and operating agencies; \(cost\) agency represents the operation cost of modular autonomous driving buses; \(cost\) passenger represents the waiting time cost of passengers, etc.; \(f\) i represents the total energy cost of vehicles with a formation number of \(i\) for a single departure; \(x\) ij is a binary variable, indicating whether to dispatch a modular autonomous driving bus with a formation number of \(i\) at time \(j\). If dispatched, then \(x\) ij =1, otherwise \(x\) ij =0; \(\delta\) represents the fixed time slot length; \(w\) represents the average waiting cost per passenger per unit time; \(q\) j-1 represents the number of passengers waiting to board at time \(j - 1\); \(a\) j represents the number of people arriving at the station on the remote parking lot side within the time slot \([j - 1, j]\); \(a\) 0 represents the initial number of waiting passengers; \(C\) F is the fixed energy cost for a single departure, \(C\) Vis the energy cost coefficient for the number of vehicles in a single formation, μ ≤ 1 is the energy calculation coefficient, c i represents the total passenger capacity of the formation with i vehicles, obtained by multiplying the number of formation vehicles i by the passenger capacity of a single modular autonomous bus;

[0078] The safety constraint is used to ensure that at most one formation of modular autonomous buses departs within a time slot. The formula is:

[0079]

[0080] The passenger number conservation constraint is used to ensure the rationality of the scheduling plan calculation. Except for the number of passengers waiting at the beginning departure time q 0 and the number of passengers waiting at the end departure time q J , the number of passengers waiting at the end of the remaining time slots is the number of passengers waiting at the end of the previous time slot, plus the number of newly arrived passengers at the remote parking lot side station, and then subtract the maximum value of the passenger capacity of the modular autonomous bus departing and 0. The formula is:

[0081] q 0 = a 0

[0082]

[0083] q J = 0

[0084] The operation quality constraint includes the overcrowding constraint and the undercrowding constraint;

[0085] The overcrowding constraint is used to ensure that the number of passengers waiting in each time slot does not exceed the maximum number of passengers Q that can be carried, in order to limit the longest waiting time of passengers. The specified time slot is 5 minutes, the maximum number of formations is 25, and the passenger capacity of each carriage is 6 people. To prevent passengers from still not being able to board the bus within 15 minutes, the maximum number of people that the system can carry is 15 / 5·25·6, that is, 450 people. The formula is:

[0086]

[0087] The undercrowding constraint is used to ensure the departure frequency. A bus must depart every three time slots. The formula is:

[0088]

[0089] In the formula, x ij-1 , x ij-2 are both binary variables, indicating whether a modular autonomous bus with formation number i is dispatched at time j - 1 and j - 2. If dispatched, it takes 1, otherwise it takes 0.

[0090] The model of the scheduling scheme generation module is solved by direct solution method, linear solution method or heuristic algorithm. The one-hour modular bus departure plan for a certain day at Guangzhou Baiyun Airport is solved by Gurobi commercial solver, as shown in Table 2.

[0091] Table 2: Estimated number of passengers arriving at the station and modular bus departure plan

[0092]

[0093]

[0094] The operation module includes a reservation mini-program, a station on the remote parking lot side, a station on the airport terminal side, modular buses, and a modular bus lane between the remote parking lot and the terminal;

[0095] The reservation mini-program is used to collect the name of the remote parking lot, flight number, boarding gate, number of passengers, arrival date and specific time, and pay the parking fee online;

[0096] Both the station on the remote parking lot side and the station on the airport terminal side include a formation line and a storage line for modular autonomous buses, a passenger waiting area, a drop-off area, and a boarding bus toll office; The station on the remote parking lot side and the station on the airport terminal side also include corresponding parking services and terminal passenger services;

[0097] The modular bus lane between the remote parking lot and the terminal is set as a dedicated lane;

[0098] The modular buses depart in formation according to the generated modular bus scheduling scheme, shuttling between the station on the remote parking lot side and the station on the airport terminal side. For multiple remote parking lots and multiple terminals, modular buses depart from different remote parking lots and merge at the route intersection; When going to different terminals, they split at the route bifurcation to complete the passenger connection between the remote parking lot and the airport terminal; As Figure 3 shown, the number of modular bus formations departing from Remote Parking Lot 1 is 1, and the number of modular bus formations departing from Remote Parking Lot 2 is 2. The two merge and drive on the overlapping section of the route. Passengers adjust the modular bus compartments they are in according to their destinations, and then the modular buses separate at the route bifurcation and drive to Terminal 1 and Terminal 2 respectively.

[0099] As Figure 4 shown, the system operation process from the perspective of passengers includes the following steps:

[0100] Passengers fill in information and pay fees on the reservation mini-program;

[0101] Passengers drive to the remote parking lot. The parking lot scans the vehicle license plate and matches it with the background reservation data; Passengers look for a parking space and park the car;

[0102] If a passenger has no checked luggage, they can directly print their boarding pass at the self-service machine on the side of the remote parking lot near the station; if a passenger has checked luggage, they should go to the terminal building first to handle the check-in and print the boarding pass.

[0103] Passengers wait in the waiting area for the bus to depart, and then take a modular autonomous bus to the corresponding terminal building via a dedicated lane.

[0104] Passengers get off at the passenger drop-off area of the terminal building and can either directly go through security or handle the check-in; the modular bus picks up departing passengers and returns to the remote parking lot.

[0105] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be equivalent replacement methods and are all included in the protection scope of the present invention.

Claims

1. An airport passenger evacuation system based on modular autonomous driving bus connection, characterized in that, it includes: A data acquisition module, which is used to fuse multi-source data to predict the number of people arriving at the remote parking lot side station in each time slot, and determine the performance parameters of the modular autonomous driving bus, the passenger time cost parameters and the operation parameters; A scheduling plan generation module, based on the constructed scheduling model, inputs the parameters obtained by the data acquisition module, and solves the departure plan of the modular autonomous driving bus, that is, when to depart and the number of vehicles in the departure formation, with the goal of minimizing the total economic cost of passengers and the operation agency; An operation module, which is used to automatically form and depart according to the departure plan of the modular autonomous driving bus, and complete the passenger connection between the remote parking lot and the airport terminal; In the scheduling plan generation module, the scheduling model is applied to the passenger connection system based on the modular autonomous driving bus with a single starting and ending point, and solves the departure plan of the modular autonomous driving bus at the starting point, that is, when to depart and the number of vehicles in the departure formation, with the goal of minimizing the total economic cost of passengers and the operation agency, including the following assumptions: Regulations For the set of vehicles, For the maximum vehicle formation number, For the vehicle formation index, ; Adopt a discretization idea to divide the time range of the departure plan to be solved into integer time slots ; Specify as the set of discrete time points, as the total number of time slots, as the discrete time point index, , Take to represent the last discrete time point within the said time range; Decision variable is a binary variable, indicating whether to dispatch a formation at time , that is, to determine whether to depart and the number of vehicles in the corresponding departure formation at each discrete time point ; The scheduling model includes an objective function, safety constraints, passenger number conservation constraints and operation quality constraints; The calculation formula of the objective function is: ； ； ； In the formula, represents the total economic cost of passengers and the operating agency; represents the operating cost of modular autonomous driving buses; represents the waiting time cost of passengers; represents the total vehicle energy cost when the number of vehicles in a single departure is ; is a binary variable, indicating whether a modular autonomous driving bus with a formation number of is dispatched at time . If it is dispatched, then , otherwise represents the fixed time slot length; represents the average waiting cost per passenger per unit time; represents the number of passengers waiting to board at time -1; represents the number of people arriving at the station on the remote parking lot side within the time slot; represents the initial number of waiting passengers; is the fixed energy cost for a single departure, is the energy cost coefficient for the number of vehicles in a single formation, is the energy calculation coefficient, represents the total passenger capacity of a formation with the number of vehicles , which is obtained by multiplying the number of formation vehicles by the passenger capacity of a single modular autonomous driving bus; The safety constraint is used to ensure that at most one formation of the modular autonomous driving bus departs within one time slot, and the formula is: ； The number conservation constraint is used to ensure the rationality of the calculation of the scheduling plan. Except for the number of people waiting at the departure time and the number of people waiting at the end time of the last bus , the number of people waiting at the end of the remaining time slots is the number of people waiting at the end of the previous time slot, plus the number of people newly arriving at the station on the remote parking lot side, and then subtracting the maximum value of the passenger capacity of the modular autonomous driving bus departing and 0. The formula is as follows; ； ； ； The operation quality constraints include overcrowding constraints and undercrowding constraints; The overcrowding constraint is used to ensure that the number of passengers waiting for the bus in each time slot is not greater than the maximum number of passengers Q that can be carried, so as to limit the longest waiting time of passengers, and the formula is: ； The undercrowding constraint is used to ensure the departure frequency, and a bus must depart every three time slots, and the formula is: ； Wherein, and are both binary variables, indicating whether a modular autonomous driving bus with a formation number of and is dispatched at time . If it is dispatched, take , otherwise take .

2. The airport passenger evacuation system based on modular autonomous driving bus connection according to claim 1, characterized in that: In the data acquisition module, the multi-source data includes mini-program reservation data and airport flight data. It is stipulated that is a set of discrete time points, is the total number of time slots, is the discrete time point index, , takes indicating the last discrete time point. The calculation formula for the number of people arriving at the remote parking lot side station in each time slot is as follows: ； In the formula, represents the estimated number of passengers arriving at the station within the time slot; represents the estimated number of passengers arriving at the station within the time slot obtained through the reservation data of the mini-program; represents the estimated number of passengers arriving at the station within the time slot obtained through the airport flight data; represents the weight, and ; The estimated number of passengers arriving at the station in each time slot obtained by booking data through a small program can be statistically obtained based on the arrival time at the parking lot filled in by users on the remote parking lot reservation small program; The estimated number of passengers arriving at the station in each time slot obtained by airport flight data, the calculation formula is: ； In the formula, represents the number of passengers arriving at the terminal within a certain period in the future after the time slot, which can be obtained based on the airport flight schedule, the passenger capacity of each flight, and the reserved occupancy rate; represents the sharing ratio of passengers choosing to take the modular autonomous driving bus to the terminal, which can be obtained based on historical operation data and research actual situation data; The function fits the distribution of passengers arriving at the remote parking lot side station over time, and projects the number of people arriving at the terminal by the modular autonomous driving bus within a certain period in the future to the time slot to obtain the number of passengers arriving at the remote parking lot side station within the The weight can be obtained based on historical data, adopting various optimization methods, with the goal of minimizing the error between the calculated value and the actual value, and it needs to be updated periodically.

3. The airport passenger evacuation system based on modular autonomous driving bus connection according to claim 1, characterized in that: In the data acquisition module, the performance parameters of the modular autonomous driving bus include the passenger capacity of a single modular vehicle, the fixed energy cost of a single departure, the energy cost coefficient of the number of vehicles in a single formation and the energy calculation coefficient; the passenger time cost parameters include the waiting cost of each passenger per unit time; the operation parameters include the minimum departure interval and the maximum number of vehicle formations.

4. The airport passenger evacuation system based on modular autonomous driving bus connection according to claim 1, characterized in that: The model solution of the scheduling plan generation module adopts the direct solution method, the linear solution method or the heuristic algorithm.

5. An airport passenger evacuation system based on modular autonomous driving bus connection according to claim 1, characterized in that: the operation module includes a reservation mini-program, a remote parking lot side station, an airport terminal side station, modular autonomous driving buses, and a modular autonomous driving bus lane between the remote parking lot and the terminal; the reservation mini-program is used to collect the name of the remote parking lot, flight number, boarding gate, number of passengers, arrival date, and specific time, and pay the parking fee online; both the remote parking lot side station and the airport terminal side station include a formation line and a parking line for modular autonomous driving buses, a passenger waiting area, a drop-off area, and a boarding bus toll office; the remote parking lot side station and the airport terminal side station also include corresponding parking services and terminal passenger services; the modular autonomous driving bus lane between the remote parking lot and the terminal is set as a dedicated lane; the modular autonomous driving buses are grouped and dispatched according to the generated modular autonomous driving bus scheduling plan, shuttling between the remote parking lot side station and the airport terminal side station; for the case of multiple remote parking lots and multiple terminals, the modular autonomous driving buses depart from different remote parking lots and merge at the route intersection; when going to different terminals, they split at the route bifurcation to complete the passenger connection between the remote parking lot and the airport terminal.

Images