Optimal scheduling of aircraft in the event of weather disruption

Sivabalan Panneer Selvam, Purdue University

Abstract

Weather is a major limiting factor in National Airspace System today and accounts for more than half of all the delays. Because we cannot control weather, our ability to predict its influence and make optimal decisions is critical in minimizing its impact on the air transportation system. In this thesis a model is presented to optimally reroute and schedule a set of aircraft which have been disrupted by a weather condition. A dynamic programming approach to determine the shortest path around a moving weather condition, implemented using MATLAB by Professor Gary Slater, University of Cincinnati was modified to determine the minimum flight time around a dynamic weather condition. The delay incurred in flying around the weather condition is added as a constraint to a mixed 0-1 integer model originally developed by Narendra Sharma, Purdue University to determine the optimal schedule. The schedule drawn using the current approach of rerouting aircraft in the shortest time around the weather condition and implementing the frozen schedule, is then compared to the optimal schedule given by the model developed. It was found that when controllers had the capability of long-term accurate weather prediction, the current approach was found to be optimal in a majority of the cases tested. However in the absence of the capability or in case of sudden weather changes, the model presented provided better schedules with a lower overall delay when compared to the current approach. A logistic regression analysis was performed to determine the factors affecting the optimality of the rerouting decisions.

Degree

M.S.I.E.

Advisors

Uhan, Purdue University.

Subject Area

Aerospace engineering|Industrial engineering

COinS