Hub-spoke Airline Network Optimization For Air-high Speed Rail Intermodal Transportation

The optimization model of the hub-spoke airline network is at a relatively simplified stage.As the scale of the civil aviation industry continues to increase,the passenger flow at the airports has increased dramatically,business processes have become more complicated,and the congestion of the hubs has also become serious.The existing optimization model is based on a strict hub-spoke network structure and does not consider the hub airport capacity constraints.As a result,the hub airport cannot carry huge passenger traffic and a large hub-spoke network,flight frequency,fewer airlines,etc.To this end,this article aims to alleviate the congestion problem at the hub airport,and carries out research on optimization of the network structure and model optimization of the hub-spoke airline network.Aiming at the existing optimization model based on the strict hub-spoke airline network structure and the constraints related to the congestion problem,a research on the hub-spoke airline network optimization model for the congestion problem is carried out.First,a network structure of hub-spoke airline including direct flights is proposed to allow direct flights between non-hub airports that partially meet the constraints.Then,the optimization model considers the change of congestion cost function,congestion constraints,and direct flight constraints brought by direct flights.In addition,the biggest problem that the direct navigation brings in the model solution is that the model optimization involves complex variables that include two transportation modes and four city node transportation routes.This model proposes a list-based network coding method combined with a route path selection strategy to form a particle initialization generation algorithm,which reduces the access space and time complexity of model variables and makes the particle initialization faster.Finally,the simulated annealing particle swarm optimization algorithm is used to reduce the probability of being trapped in a local optimal solution and increase the global search efficiency of a larger solution space.Validated on the internationally publicized Turkish traffic data set,the experimental results show that compared with the strict hub-spoke airline network,the non-strict hub-spoke airline network can significantly ease the congestion of the hub airport,balance the passenger traffic between the hub airports,and reduce network costs;The proposed algorithm in this model has fast convergence speed and good stability.In order to further solve the congestion problem at the hub airport,and to meet the needs of comprehensive transportation development,the high-speed rail station is considered in the hub-spoke airline network,and an optimization model of the hub-spoke airline network for air-rail intermodal transportation is proposed.First,the optimization model adds a high-speed rail node to the non-strict network structure of the hub-spoke airline network,and determines the current node's transportation mode and whether it is a hub node based on factors such as distance and cost.Then,because the high-speed rail nodes are considered,the model variable represents a transportation line that includes three transportation modes and four city nodes.The access complexity of this variable is more complicated than the model variable in the hub-spoke airline network optimization model for the congestion problem.Therefore,a high-dimensional multi-angle scale variable representation method and a Tabu search algorithm solution method are designed.In the end,this method greatly reduces the computational burden of the algorithm,and reasonably distributes the mode of transportation and passenger flow on the line.In order to further verify the universality of the network structure of the hub-spoke network,the validation experiments on the domestic transportation data set shows that the optimization model can significantly alleviate congestion at hub airports,equalize passenger traffic between hub airports,and reduce network costs compared to strict and non-rigid hub-spoke airline network.At the same time,it proves that the model solving method has good stability.