Dynamic Real-time Scheduling of Terminal Traffic

摘要

This paper presents dynamic strategies for the integrated scheduling and runway assignment of both arrival and departure traffic over an airport. While a static scheduling scheme handles traffic over a specified planning horizon simultaneously, sequential dynamic schedulers divide the planning horizon into a series of smaller scheduling windows and apply a static scheduling scheme sequentially over each window. Dynamic scheduling strategies are desirable for obtaining real-time solutions of continual traffic streams and for taking advantage of updated traffic information. In this paper, a multiple-point scheduling framework is used in which scheduling locations include runway thresholds as well as fixes over the terminal airspace and gates on the airport surface. Integrated static scheduling of both arrival and departure traffic is formulated as mixed-integer linear programming (MILP). Solution variables include scheduled times of arrival (STA) at the multiple scheduling locations and aircraft sequences at merge points. In addition, aircraft route assignments for both ground and airborne traffic are included as discrete solution variables, from which optimal runway assignments can be determined. Then, different dynamic strategies with either overlapping or non-overlapping scheduling windows are developed and compared. Induced constraints for ensuring sufficient separations among traffic in neighboring windows are discussed. Real traffic data from the JFK airport is used in extensive numerical solutions to evaluate the computational speeds and scheduling performances of different dynamic strategies.