Airline schedules have traditionally been set long in advance, and any deviations from the present schedule have been due to weather or maintenance interruptions. A static schedule of this type cannot respond to the variations in demand that inevitably occur on a daily or even weekly basis. This dissertation investigates the potential for airlines to add an element of flexibility to their schedule by dynamically reassigning different aircraft to routes. These reassignments must not violate the unique constraints an airline is faced with, such as crew, maintenance and operational issues.;Aircraft are dynamically reassigned to routes through aircraft swaps, or substitutions between a pair of aircraft. Aircraft swaps are identified by using a set of algorithms that search for oversold and undersold flights in the network and then suggest feasible substitutions between aircraft. Forecast day-of-departure passenger demand is the criteria for identifying these swaps. A second set of algorithms examines each potential swap for adherence to constraints such as crew and maintenance. Swaps that do not meet any of the constraints are eliminated. The remaining swaps are checked for a forecast increase in profit if the swap is executed. A network flow model identifies the optimal swaps from the set of feasible, profit-improving swaps.;This research used the actual March, 1995, United Airlines schedule. This schedule included initial aircraft assignments as well as basic maintenance constraints and crew routings. A simulation model incorporated the actual United schedule and used generic booking data to model passenger arrivals.;Results from the model include a nominal increase of 0.54 percent in operating profit. Passenger load factor increased 0.2 percent and aircraft utilization for the smallest aircraft types increased. Approximately 5 percent of the daily schedule was affected by aircraft swaps.
展开▼
