Network Congestion Control of Airport Surface Operations

A model of airport operations using actual surface surveillance data is proposed. The study modeled the aircraft taxi-out process by representing the airport surface as a network and the total taxi time as the sum of the travel times along different links in the surface trajectory of an aircraft. A set of suitable random processes was proposed to model the distribution of link travel times. A control algorithm that attempts to minimize the delay cost for aircraft while limiting the effect on throughput was developed. Using the test cases of a simple single-link model as well as the complete airport network for Boston Logan International Airport, it was shown that the algorithm can successfully maintain the level of surface traffic at a specified average value. When suitably calibrated, this steady traffic level can provide significant reduction in aircraft surface fuel burn without adversely affecting airport performance.