Column Generation Technique for Crew Rostering Problem with Workload and Earnings Balancing

Abstract

Abstract

A crew rostering problem in the airline industry is the problem of constructing a crew schedule which consists of flights, pairings, training activities, and days off for each crew member under some certain constraints; cost minimization is also an important concern that must be considered. The size of this problem is usually huge since the number of flights, pairings and crews are large; the constraints are also generally complex. In this paper, we present an approach for solving a crew rostering problem whose objective is to balance the workload and earnings among crew members. The problem is modeled as a set partitioning problem and solved using a column generation technique. Some numerical results are presented and analyzed using the data from Thai Airways International as a case study.

Keywords : Column Generation; Crew Rostering; Set Partitioning; Workload Balancing