Analysis of three mathematical models of the Staff Rostering Problem

Staff rostering is a major challenge in the service sector, where exploitation costs are essentially made up of staffing costs. Searching for an optimum has direct economic returns but the rosters must satisfy numerous legal constraints. This paper presents work on an exact approach using branch-and-price methods on a concrete situation. We develop three MILP models and extend them with valid inequalities to two cases. Their computation results on a set of 960 tests covering several scenarios will then be compared and analyzed.