Search

CN-121998345-A - Airport ground guarantee vehicle cooperative scheduling optimization method and system

CN121998345ACN 121998345 ACN121998345 ACN 121998345ACN-121998345-A

Abstract

The invention discloses an airport ground guarantee vehicle collaborative scheduling optimization method and system, which belong to the technical field of airport ground guarantee vehicle collaborative scheduling, and comprise the steps of S1, selecting a model according to an optimization target, S2, aiming at the selection model, carrying out model solving, generating a feasible initial solution, relaxing vehicle constraint, setting a large virtual vehicle upper limit for each vehicle team to ensure that a solution space is feasible, changing an optimization target, temporarily minimizing the total number of vehicles of the vehicle team, running ALNS, carrying out iterative optimization by using a damage operator and a repair operator, reducing the number of vehicles, stopping when the number of vehicles of each vehicle team does not exceed the actual number, obtaining the feasible solution, carrying out target driving optimization, carrying out optimization by using the feasible solution as a current solution, carrying out operation enhancement ALNS after recovering an original target, carrying out self-adaptive damage and repair operation, carrying out iterative execution on the damage and repair operation, receiving a new solution based on a simulated annealing criterion, dynamically updating operator weight, and finally outputting an optimal scheduling scheme, and outputting and executing the optimal scheduling scheme.

Inventors

  • ZHAO YONGKANG
  • Ding Chenghong
  • FENG XIA

Assignees

  • 中国民航大学

Dates

Publication Date
20260508
Application Date
20260127

Claims (10)

  1. 1. The airport ground guarantee vehicle cooperative scheduling optimization method is characterized by comprising the following steps of: S1, selecting a model according to an optimization target: when the optimization target is the shortest total distance travelled by the vehicle, selecting a travelling distance optimization scheduling model; when the optimization target is that the total delay time is minimum, selecting a delay time optimization scheduling model; Selecting a pareto multi-target optimization model when the optimization targets merge the total driving distance and total delay time of the vehicle; S2, aiming at the selected model, carrying out model solving, including: s201, generating a feasible initial solution, namely relaxing vehicle constraint, setting a large virtual vehicle upper limit for each vehicle fleet k to ensure that a solution space is feasible, changing an optimization target, temporarily minimizing the total vehicle number of the vehicle fleet, running ALNS, performing iterative optimization by using a destruction operator and a repair operator, reducing the vehicle number, and ending when the vehicle number of each vehicle fleet does not exceed the actual number to obtain the feasible solution; S202, target driving optimization, namely taking a feasible solution as a current solution, carrying out optimization by running enhancement ALNS after recovering an original target, carrying out self-adaptive selection of a damage and repair operator, carrying out iterative execution of the damage and repair operation, receiving a new solution based on a simulated annealing criterion, and dynamically updating the weight of the operator; s3, outputting and executing the optimal scheduling scheme.
  2. 2. The airport surface assurance vehicle co-scheduling optimization method of claim 1, wherein the mathematical expression of the travel distance optimization scheduling model is: (1) (2) (3) (4) (5) (6) (7) Wherein: Representing the driving cost of a vehicle team k on an edge (i, j) as a cost function, wherein i and j are the position nodes where flights are located; as a decision variable, if the vehicle v in the vehicle team k runs from the machine position node i to the machine position node j, the vehicle v is 1, otherwise, the vehicle v is 0; Is a collection of fleet k; representing a certain vehicle in the fleet k; Is a collection of garage and machine position nodes; is a set of machine node; For decision variables, if a vehicle v in the fleet k travels from the level node i to the level node Then 1, otherwise 0; for decision variables, if the vehicle v in the vehicle fleet k is from the station node The running to the machine node j is 1, otherwise, the running to the machine node j is 0; For decision variables, if vehicle v in fleet k is from garage node The running to the machine node j is 1, otherwise, the running to the machine node j is 0; For decision variables, if a vehicle v in the fleet k travels from the level node i to the garage node Then 1, otherwise 0; As a real decision variable, representing the start service time of the vehicle v in the fleet k to the flight on the position node j; As a real decision variable, representing the start service time of the vehicle v in the fleet k to the flight on the position node i; The service duration for fleet k for flights on location node i; the time for fleet k to travel from node i to node j; the maximum time allowed to stay on the fleet after servicing the vehicles v in fleet k; The demand of the flight on the machine position node i for the resource; the maximum amount of resources carried for the vehicles in fleet k; The service duration for fleet k for flights on location node i; the time for fleet k to travel from node i to node j; the lower bound of the time when the vehicle in the motorcade k reaches the machine node i; For real decision variables, representing fleet The vehicle v in (1) serves the start service time of the flight on the location node i; For motorcades Service duration for flights on the level node i; For real decision variables, representing fleet The vehicle v in (1) serves the start service time of the flight on the location node i; The upper bound for the time that the vehicle in fleet k arrives at level node i.
  3. 3. The airport surface assurance vehicle co-scheduling optimization method of claim 2, wherein the mathematical expression of the delay time optimization scheduling model comprises: Calculating the completion time of each service: (13) (14) calculating delay time: Introducing auxiliary variables Represents the latest completion time for flight i: Introducing binary variables When the flight i delays Otherwise ; The objective function is: , Wherein: Is that ; Is that ; Is that ; A real decision variable, representing the start service time of a vehicle v in a catering fleet to a flight on a machine node i; service duration for the catering service for flights on the location node i; A real decision variable, representing the start service time of a vehicle v in a freight train to a flight on a position node i; service duration for the cargo service to the flight on the location node i; For a real decision variable, representing the start service time of a vehicle v in a ferry fleet to a flight on a machine node i; service duration for the ferry service for the flight on the level node i; Is the machine position ; M is ; Is that ; Objective function: , Preserving constraints (2) - (12); Newly adding constraint conditions (13) - (21); by constraining (16) - (18) Equal to the maximum of three service completion times, ensuring delay times by constraints (19) - (21) When (1) Indicating no delay when In the time-course of which the first and second contact surfaces, Indicating a delay time.
  4. 4. The airport surface assurance vehicle co-scheduling optimization method of claim 1, wherein the disruption operators comprise random removal, shaw removal, worst removal, path removal, cluster removal, time window removal, and demand removal.
  5. 5. The airport ground assurance vehicle co-scheduling optimization method of claim 1, wherein the repair operator comprises a greedy repair algorithm and a regret repair algorithm.
  6. 6. An airport ground assurance vehicle co-scheduling optimization system, comprising: the man-machine interaction module is used for inputting basic data, selecting a model and displaying an optimization result; The model database is used for storing a running distance optimization scheduling model based on the shortest running total distance of the vehicle, a delay time optimization scheduling model based on the smallest total delay time and a pareto multi-target optimization model fusing the running total distance of the vehicle and the total delay time; the scheduling optimization module is used for importing the basic data into a selection model to carry out model solving and comprises the following steps: The method comprises the steps of generating a feasible initial solution, relaxing vehicle constraint, setting a large virtual vehicle upper limit for each vehicle fleet k to ensure that a solution space is feasible, changing an optimization target, temporarily minimizing the total vehicle number of the vehicle fleet, running ALNS, performing iterative optimization by using a damage operator and a repair operator, reducing the vehicle number, and ending when the vehicle number of each vehicle fleet does not exceed an actual number to obtain the feasible solution; And the target driving optimization comprises the steps of taking a feasible solution as a current solution, carrying out operation enhancement ALNS for optimization after recovering an original target, adaptively selecting a damage and repair operator, carrying out iterative execution of the damage and repair operation, receiving a new solution based on a simulated annealing criterion, dynamically updating the weight of the operator, and finally outputting an optimal scheduling scheme.
  7. 7. The airport surface assurance vehicle co-scheduling optimization system of claim 6, wherein the mathematical expression of the distance-of-travel optimization scheduling model is: (1) (2) (3) (4) (5) (6) (7) Wherein: Representing the driving cost of a vehicle team k on an edge (i, j) as a cost function, wherein i and j are the position nodes where flights are located; as a decision variable, if the vehicle v in the vehicle team k runs from the machine position node i to the machine position node j, the vehicle v is 1, otherwise, the vehicle v is 0; Is a collection of fleet k; representing a certain vehicle in the fleet k; Is a collection of garage and machine position nodes; is a set of machine node; For decision variables, if a vehicle v in the fleet k travels from the level node i to the level node Then 1, otherwise 0; for decision variables, if the vehicle v in the vehicle fleet k is from the station node The running to the machine node j is 1, otherwise, the running to the machine node j is 0; For decision variables, if vehicle v in fleet k is from garage node The running to the machine node j is 1, otherwise, the running to the machine node j is 0; For decision variables, if a vehicle v in the fleet k travels from the level node i to the garage node Then 1, otherwise 0; As a real decision variable, representing the start service time of the vehicle v in the fleet k to the flight on the position node j; As a real decision variable, representing the start service time of the vehicle v in the fleet k to the flight on the position node i; The service duration for fleet k for flights on location node i; the time for fleet k to travel from node i to node j; the maximum time allowed to stay on the fleet after servicing the vehicles v in fleet k; The demand of the flight on the machine position node i for the resource; the maximum amount of resources carried for the vehicles in fleet k; The service duration for fleet k for flights on location node i; the time for fleet k to travel from node i to node j; the lower bound of the time when the vehicle in the motorcade k reaches the machine node i; For real decision variables, representing fleet The vehicle v in (1) serves the start service time of the flight on the location node i; For motorcades Service duration for flights on the level node i; For real decision variables, representing fleet The vehicle v in (1) serves the start service time of the flight on the location node i; The upper bound for the time that the vehicle in fleet k arrives at level node i.
  8. 8. The airport surface assurance vehicle co-scheduling optimization system of claim 7, wherein the mathematical expression of the delay time optimization scheduling model comprises: Calculating the completion time of each service: (13) (14) calculating delay time: Introducing auxiliary variables Represents the latest completion time for flight i: Introducing binary variables When the flight i delays Otherwise ; The objective function is: , Wherein: Is that ; Is that ; Is that ; A real decision variable, representing the start service time of a vehicle v in a catering fleet to a flight on a machine node i; service duration for the catering service for flights on the location node i; A real decision variable, representing the start service time of a vehicle v in a freight train to a flight on a position node i; service duration for the cargo service to the flight on the location node i; For a real decision variable, representing the start service time of a vehicle v in a ferry fleet to a flight on a machine node i; service duration for the ferry service for the flight on the level node i; Is the machine position ; M is ; Is that ; Objective function: , Preserving constraints (2) - (12); Newly adding constraint conditions (13) - (21); by constraining (16) - (18) Equal to the maximum of three service completion times, ensuring delay times by constraints (19) - (21) When (1) Indicating no delay when In the time-course of which the first and second contact surfaces, Indicating a delay time.
  9. 9. A computer program product comprising a computer program which when executed by a processor implements the airport ground assurance vehicle co-scheduling optimization method of any one of claims 1-5.
  10. 10. A computer readable storage medium storing a computer program which when executed by a processor implements the airport ground assurance vehicle co-scheduling optimization method of any one of claims 1-5.

Description

Airport ground guarantee vehicle cooperative scheduling optimization method and system Technical Field The invention belongs to the technical field of cooperative scheduling of airport ground assurance vehicles, and particularly relates to a cooperative scheduling optimization method and system for airport ground assurance vehicles. Background With the continuous rapid development of the global civil aviation industry and the remarkable return of the air transportation demand, the complexity and the operation scale of airport ground guarantee operations are rapidly expanding. The aviation traffic volume is gradually restored to the pre-epidemic level and keeps steadily growing, so that airport ground guarantee tasks become increasingly heavy under the current operation environment, and higher requirements are put on the scheduling efficiency, the instantaneity and the synergy. The ground guarantee operation covers a plurality of key links such as oiling, cabin cleaning, catering, baggage and freight processing, passenger ferry and the like, and a plurality of types of special vehicles are required to be mobilized to work cooperatively, strict time sequence logic and space constraint exist between each operation, and the scheduling complexity is extremely high. At present, among the factors influencing the flight arrival rate, besides the uncontrollable weather reasons, delays related to airport ground guarantee operation, including long waiting time of guaranteeing vehicle service, unsmooth resource allocation, multiple service disconnection and the like, become one of the most main controllable factors causing the flight delay. This phenomenon directly indicates that a scheduling pattern lacking systemicity and synergy can significantly compromise overall operating efficiency, even directly translate into flight delays. In this context, how to construct a ground vehicle scheduling system that is scientific, efficient, real-time, and capable of coordinating multiple types of resources has become a core challenge in improving airport operation performance and management level. In the field of airport ground guarantee vehicle dispatching, the most similar prior art scheme of the invention mainly comprises the following categories, and each category has inherent serious defects to be overcome: Based on a heuristic algorithm (such as a genetic algorithm and a traditional vehicle path problem solver) of single-job optimization, the scheme describes that the method generally models the scheduling problem of a certain type of guarantee vehicles (such as only for fuelling vehicles or only for ferrying vehicles) in an airport as a classical vehicle path problem (VehicleRoutingProblem, VRP) or a common variant (such as VRP with a time window) independently, and solves the problem by adopting a genetic algorithm, simulated annealing, tabu search and other meta-heuristic algorithms. The optimization targets are often single, and the shortest total travel distance or the minimum task completion time of the vehicles are usually used as the main optimization direction. The method has the defects of isolation optimization and serious lack of cooperativity, and the scheduling optimization of a single type of vehicle is considered only from a local angle, so that the problems of strict time sequence dependency relationship (for example, passenger cabin cleaning can be carried out after passenger unloading and baggage unloading are finished, meal distribution can be carried out after cleaning is finished, aircraft refueling is carried out after most ground operations are finished and before pushing out) and space resource contention in a shared area are completely ignored. The local optimization mode often causes scheduling plan conflict among subsystems, but not the global optimization, even an actual executable collaborative operation scheme is difficult to form, so that the problems of disordered service sequence, mutual blocking of vehicles, delayed critical tasks and the like are caused in actual operation, and finally flight delay is caused. A method based on a multi-objective mathematical programming model and an accurate solver (such as CPLEX and Gurobi) is characterized in that part of research attempts to integrate scheduling requirements by constructing a Mixed Integer Linear Programming (MILP) model containing multiple types of vehicles and multiple constraints, and solving by means of the CPLEX, gurobi and other commercial mathematical programming solvers. Such models may attempt to mathematically achieve comprehensive optimization while simultaneously considering multiple objectives such as vehicle driving costs, total service time, partial resource occupancy, etc. The method has the defects that the calculation complexity grows exponentially and is difficult to be used for actual real-time scheduling, airport ground guarantee scheduling belongs to a typical NPhard problem, the scale of the problem expands sharp