Combining VNDS with Soft Global Constraints Filtering for Solving NRPs

Jean-Philippe Metivier (1) , Patrice Boizumault (1) , Samir Loudni (1)

1 Equipe CODAG - Laboratoire GREYC - UMR6072

Nurse Rostering Problems (NRPs) consist of generating rosters where re- quired shifts are assigned to nurses over a scheduling period satisfying a number of constraints. In [25], we have shown how soft global constraints can be used to model NRPs in a concise and elegant way. In this paper we go one step further by proposing new neighborhood heuristics for VNS/LDS+CP. Experiments show that, despite its genericity and exibility, our approach supplies excellent results on small and middle size problems and very promising results on large scale problems.

Mots clés: NRP; Constraint Programming; VNS; VNDS; Soft Global Constraints

Domaines: Intelligence artificielle [cs.AI]; Traitement du texte et du document

Jean-Philippe Metivier, Patrice Boizumault, Samir Loudni. Combining VNDS with Soft Global Constraints Filtering for Solving NRPs. 8th Conference on the Practice and Theory of Automated Timetabling (PATAT'10), Aug 2010, Belfast, Ireland. 14 p. ⟨hal-01017218⟩