Methods and measurements in real-time air traffic control system simulation part 2
Edward P. Buckley
B. Deano DeBaryshe
Buckley, E.P., DeBaryshe, B.D., Hitchner, N., & Kohn, P. (1983). Methods and measurements in real-time air traffic control system simulation (DOT/FAA/CT-83/26). Atlantic City International Airport: Federal Aviation Administration Technical Center.
The major purpose of this work was to assess dynamic simulation of air traffic control systems as a technique for evaluating such systems in a statistically sound and objective manner. A large set of customarily used measures based on the system mission of safe and expeditious movement of air traffic was collected by the computer generating the simulated traffic. The measures were collected running 1-hour simulation exercises. These measures were applied in two experiments .involving controllers performing traffic control in single en route sectors, with coordination with simulated adjacent sectors. Two experiments, having many replications, were conducted. In addition to studying the characteristics of the set of measurements, a second aim of. the first experiment was to determine the effect on the measurement of surrounding circumstances, specifically sector geometry and traffic density. The results of this experiment led to a decision to conduct a much less complex experiment, confined to only one sector and geometry but with more repetitions of l-hour runs under the same circumstances. This enabled an examination of the use of aggregation of data to improve reliability and the execution of a factor analysis in order to reduce and simplify the set of measures.
The factor analysis reduced the measure set to four factor scores. The study of aggregation of data led to the conclusion that four hours should be the minimum data point basis. The data from the first experiment was then utilized to cross validate the four factor structure, which had been found. This cross validation was reasonably successful. The use of the four factor scores and their primary original scores, plus two auxiliary measures, is recommended in future air traffic control system dynamic simulations for system test and evaluation.
Updated: May 04, 2012 11:21 AM