Research On Key Techniques Of Air Traffic Flow Management In Airport Terminal Area Based On Uncertain Capacity Conditions

The change of capacity is a significant factor that may cause flight delay. Especially for airportterminal area with large number of flights and heavy workload of air traffic management. Theuncertainty of airport capacity resulted from weather brings much difficulty in terminal air traffic flowmanagement. Airport terminal area has been a bottleneck for domestic or world-wide air trafficmanagement. Studying the key technologies of terminal air traffic flow management which adapt touncertain capacity conditions is with important theoretical and practical significance for ensuring thestability of air traffic system, improving the efficiency of air traffic flow management and reducingflight delay.On the basis of literature review for domestic and foreign related research, the paper studied severalkey problems of terminal air traffic flow management with uncertain capacity conditions. The maincontents and achievements are listed below.First, to solve collaborative slot allocation with airport probability capacity, a set of stochasticcollaborative slot allocation models was proposed. With the aim to minimize the expected total flightdelay cost and the expected average delay time of passengers, the principles of efficiency and equitywere established. On the basis of the constraints of effectiveness, slots were allocated to arrival flights.According to the form of airport capacity being updated, both static model and dynamic model forwere proposed. Non-dominated sorting genetic algorithm was applied in solving the models, andstatic stochastic RBS and dynamic stochastic RBS were proposed based on tradditional RBS. Set theflight schedul of one of domestic airports as an example, a numerical test was performed. Test resultsshow that, compared with typical stochastic models and stochastic RBS algorithms, the modelsproposed win them and could efficiently, equitably and effectively allocate slot with airportprobability capacity. Test results also show that, the models improve the flexibility of slot allocationand enrich the decision space.Second, to effectively perform ground delay program with airport uncertain capacity, a set of robustground delay program models was proposed. With the aim to respectively minimize the total flightdelay cost in each capacity scenario, the maximum deviation between total flight delay cost and theoptimal total flight delay cost in each capacity scenario as well as the ratio, the principles ofrobustness were established. According to the form of airport capacity being updated, both staticmodel and dynamic model for were proposed. Genetic algorithm was applied in solving the models.Set the flight schedul of one of domestic airports as an example, a numerical test was performed. Test results show that, compared with stochastic models under equal probability scenarios, the modelsproposed are better than them in achieving the goals of robustness and provide available reference fordecision making.Third, to solve collaborative slot allocation with airport uncertain capacity, a set of robustcollaborative slot allocation models was proposed. Set the total flight delay cost and the average delaytime of passengers as the basis of the principles of efficiency and equity respectively, absolute robust,deviation robust, and relative robust models were established. On the basis of the constraints ofeffectiveness, robust optimal strategies for slot allocation were searched for. According to the form ofairport capacity being updated, both static model and dynamic model were proposed. Non-dominatedsorting genetic algorithm was applied in solving the models. Set the flight schedul of one of domesticairports as an example, a numerical test was performed. Test results show that, compared withstochastic collaborative models under equal probability scenarios, the models proposed couldefficiently, equitably and effectively allocate slot and improve the stability of slot allocation withuncertain capacity.Forth, to solve the arrival and departure flow allocation with airport probability capacity, a set ofstochastic flow allocation models for both single airport terminal area and multi-airport terminal areawas proposed. The strategy for maintaining the risk of flight delay was introduced in. With the aim tominimize the total flight delay cost under certain risk of flight delay, the models utilized the capacityand optimized the arrival and departure flow. A new intelligent algorithm named predatory searchalgorithm was applied in solving the models. Set the flight schedul of one of domestic terminals as anexample, a numerical test was performed. Test results show that, compared with typical models, themodels proposed gain a better effect than them and could effectively reduce flight delay cost with thesame delay risk.Fifth, to solve the arrival and departure flow allocation with airport uncertain capacity, a set of flowallocation robust optimization models for both single airport terminal area and multi-airport terminalarea was proposed. With the aim to respectively minimize the total delay cost with certain robustfactor, average value plus standard deviation of delay cost in each capacity scenario and the maximumloss of delay cost, regreat robust, differential robust and preference robust models were proposed.Predatory search algorithm was applied in solving the models. Set the flight schedul of one ofdomestic terminals as an example, a numerical test was performed. Test results show that, comparedwith typical models, regreat robust models proposed gain a better effect than them with the samerobust factor. Test results also show that, differential robust and preference robust models proposedcould effectively be with robustness on the basis of different decision preferences and reduce thedisturbances of flight delay in each capacity scenario. The paper enriched the theory and method of terminal air traffic flow management, and provided anew way to study air traffic flow management with uncertain capacity. Besides, the paper provided areference for air traffic flow management decision making in real world.