Research On The Airport Ground Taxiway Scheduling Problem Based On Fusion Ant Colony Algorithm

With the air travel booming, the running environment of busy airports becomes more complex and uncertain, which makes the taxi scheduling plan by artificial means more difficult. Scheduling using a first-come, first-service, although it can reduce the workload of staffs, may make some sections of congestion and delays in delivery. Therefore, how to determine the flight glide order and how to reasonably determine the position of each flight and glide path, can effectively avoid taxiing conflict, which has a great impact on the results of taxiway scheduling.The main work of this paper is as follows:Firstly, the passer briefly discusses the basic knowledge of the important airport resources of taxiway, runway and apron and so on, the network map of the airport network,the mathematical model of taxi scheduling problem with the design objective function, the complexity analysis on the taxiway the scheduling problem.Secondly, this paper introduces three types of taxi conflict and researches on taxi conflict detection and relief. To cross conflicting characteristics, a conflict detection and release algorithm is designed based on a two-phase lock; by analysis of the characteristics of the different types of conflict resolution, a complete glide conflict detection and relief algorithm is designed.Finally, by analyzing the advantages and disadvantages of genetic algorithm and ant colony algorithm, a fusion ant colony algorithm is proposed. The algorithm consists of two phases, by learning the "phase integration" idea. The first stage is used for the coarse search, in order to achieve the suboptimal solution, in addition to initializing the pheromone of the ant colony algorithm and determining the glide scheduling order of ant colony algorithm; the second stage is a detailed search, in order to obtain a complete solution. By the comparison with ant colony algorithm, fusion ant colony algorithm shows achieves satisfactory results in the performance and effectiveness of algorithm.