Multiple airport ground holding dynamic model and optimal algorithm basing flight priority

With the increase of air transportation demand, the problem of air traffic congestion has become more and more serious, so the study of flow optimization management is of urgent need and great importance. Ground-holding policy (GHP) is regarded as one of the most effective methods of short-term air flow management to ease the air traffic congestion and improve the air safety. The essence of the GHP optimal algorithm is to decide which flight is to be delayed to take off and how long it is delayed. The study of GHP is concerned about single airport ground holding and multi-airport ground holding, and this paper focuses on the latter which is more complex. To present the special features of dynamic air flow management, the paper sets up a new multi-airport ground-holding dynamic model with en-route capacity constraints in sector. Because the problem includes the large number of holding aircraft, multi-airport ground holding problem itself is a NP problem which is hard to achieve a solution. The problem becomes more complex when considering some factors like capacity constraint in sector and delay cost of different aircraft. But in fact, the controller wants an instant satisfying solution, but is not an exact solution using long time. To achieve instant satisfying solution, we design a heuristic optimal algorithm basing on flight priority in the paper. Each flight, considering continuance flight, scheduled arrival time, cost of delay, is given priority. Each airport is given slot according its capacity. Then base the flight priority, make each flight as shift point, the slots have been allotted. If the flight sequence can not satisfy the capacity constraint, then adjust the arrival time of the flight according some rules until they satisfy all airport capacity constraints. To conclude, the algorithm hammers to approve a real-time solution which satisfying all capacity constraints and low delay cost. The simulation results validate the feasibility of the proposed model - - and algorithm.