Using Ant Colony Optimization Algorithm to Solve Airline Crew Scheduling Problems

• Main Authors: Raymald, 鄧廣豐
• Other Authors: Lo, Chih Chung
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/79802535705267284094

碩士 === 佛光人文社會學院 === 資訊學研究所 === 93 === Abstract Crew cost is essential to airline carrier operations. It is important for airline carriers to design good crew schedules to reduce crew cost. Since a feasible crew schedule has to satisfy a drafted flight schedule, fleet routes, agreements with labor unions, government regulations and a carrier’s own policy, it is complicated. Crew scheduling is an NP-hard constrained combinatorial optimization problem, very important for the airline industry. So this problem has been widely studied in literatures. Airline crew scheduling problems have been traditionally formulated as set covering problems or set partitioning problems. But this approach when flight networks are extended, this problem becomes more complicated and thus more difficult to solve. So we solve this problem using a algorithm applied to a flight graph representation. In this thesis, we design an Ant Colony Optimization Algorithm (ACO) for the airline crew scheduling problem. We adjust the ACO algorithm and propose Ant Colony Scheduling Model (ACSM) for the crew scheduling. Even so, our research is not only provided an efficient solution for crew scheduling problems, but also focuses on expanding the ACO application by applying it. To evaluate the performance of our method, we adopted two case study regarding the international operations of a major Taiwan airline in literature as our test problems, each has a problem size of 68 flights and 454 flights in a week respectively. Under the same scheduling constraints, computational results of the first test problem showed that ACO yields an improvement of 5.35%, 2.41% and 2.41% over the another (published best-known) three solutions in literatures, respectively, and the another showed that an improvement of 1.110% and 0.995% over the another (published best-known) two solutions in literatures, respectively. Therefore, ACSM we have proposed is an efficient solution algorithm to help carriers minimize crew cost and plan proper crew pairing for airline crew scheduling problem.