Research On Slot Allocation Models And Algorithms In Ground Holding Policy

Ground holding policy (GHP) is one of the effective methods to alleviate congestions and its core problem is slot allocation problem. In this dissertation, the history development and the the latest research results of slot allocation in GHP are summarized in detail. Besides, based on these attributes, slot allocation models and algorithms are researched for centralized and decentralized allocation patterns respectively. This provides a theoretical basis for scientific and rational allocation of slots. The foundation for achieving effectiveness is to determine the number of slots can be used. Several concepts are defined, such as, capacity scenarios clustering; nominal capacity scenario, nominal-capacity-scenarios tree, nominal capacity value, effective capacity and effective slots, and then their mutual relationships are analyzed. Then SOM neural network mixed with k-means clustering algorithm are used to cluster historical data of capacity scenarios, based on which nominal capacity value and corresponding probability are obtained. Finally, effective capacity method is researched by using stochastic GHP model and mathematical induction method, followed by examples simulation to validate above discrimination method.Aiming at single-objective centralized slot allocation, balanced allocation modes are builded adding equity. Reduction in delay time or delay cost characterizes efficiency and equity has to be based on specific issues. Therefore, efficiency and quity are quantified and then with objective function of efficiency and constrains of effectiveness and equity, single-objective optimization models are established. After that, artificial fish school algorithm (AFSA) is applied to sovle models and numeric simulations are carried out.Aiming at multi-objective centralized slot allocation, several objectives of airlines are expanded, such as improving flights on-time rate, reducing passenger delay time, and so on. Then, multi-objective optimization models are established to meet the decision-making objectives of the department of air traffic flow management and airlines. Finally, improved AFSA is applied to seek Pareto solutions for multi-objectives optimization problems and numeric simulations are carried out.Aiming at slot-exchange decentralized allocation, with the guidline of game theory, slot allocation modes and heuristic algorithms are researched. Airline's decision-making objectives characterize efficiency and equity is characterized by meeting these objectives. Under the condition of complete and perfect information, single objective and multi-objectives are researched in one game process respectively. And then, three alogorithms are proposed to allocate slots, including backward induction, fixed priority TTC and dynamic priority algorithms. After that, numeric simulations are carried out. Aiming at slot trading decentralized allocation, with the guidline of auction theory, the allocation rules and payment rules are researched. Efficiency is characterized by maximizing overall benefits and equity is characterized by obtaining the bids what they want. Under the condition of incomplete information, single auction and combinatorial auction models are established to achieve optimization allocation. Then, based on vickrey payment rule, a variety of improved payment rules are designed to meet budget balance. After that, numeric simulations are carried out to analysize the impact of these payment rules and bid policy.The dissertation is a deep research on models and algorithms of slot allocation problem in GHP. The proposed methods have strong pertinence and are easy to be realized. The research of this dissertation is conducive to improve our country's study level in this field. It not only consummate slot aloocation theory and methods, but also attributes to the further application of collaborate decision making in air traffic management.