nous n'avons résolu leprobì eme que pour un employé. Ceprobì eme ne pouvant alorsêtrealorsêtre modélisé qu'` a l'aide d'une seule contrainte multicostregular . Les r` egles 2, 3, 5 et 6 sont modéliséesmodéliséesà l'aide d'un automate fini ? s . Les autres seront modélisées modéliséesà l ,
optimisation multicostregular pouvaientêtrepouvaientêtre appliquéesappliquéesà la modélisation etàetà la résolution de PSP type Nurse Rostering Problems (NRP) sur-contraints. Nous avons devisé d'une méthode de résolutionincompì ete de type LNS entì erement automatisable, depuis la lecture de l'instancè a la sortie des solutions Notre implémentation est libre et open-source, et téléchargeablè a https://github.com/sofdem/chocoETP. Les parseurs pour différents jeux de test sont disponibles : shoe Les résultats expérimentaux sur les instances PATAT et ASAP sont présentés dans la section suivante. Des améliorations sontévidemmentsontévidemment encore envisageables. Notamment, certaines instances ne peuventêtre peuventêtre résolues efficacement si l'utilisateur ne précise pas une borne supérieure suffisamment serrée, ou la limite maximale de pénalitépénalitéà partir de laquelle une contraire est considérée dure. Dans les deux cas, il s'agirait d'intensifier la recherche de bonnespremì eres solutions avant de lancer la recherche locale. Il s'agirait par exemple, de brancher sur les coûts demanì ere destructive pour activer la back-propagation au plus tôt, Conclusion Dans ce chapitre ,
présentée dans ce chapitre s'appliqué egalementàegalementà la variante de satisfaction du pro-bì eme. Si elle n'est plus alors utilisée pour comptabiliser les coûts de violation, la contrainte multicostregular reste utile pour modéliser les compteurs d'activités ou de motifs. Nous avons présenté cette approche ,
nous avons présenté le modèle et les stratégies de recherche adaptéesadaptéesà ces contributions. Dans ce chapitre nous utilisons le cadre construit pour résoudre diversprobì emes issus d'une part de la compétition de planification dinfirmì eres NRP10 ,
heuristique basée sur les regrets qui exploite la structure de graphe des contraintes multicost-regular. 10. ´ Evaluations 10.2 ASAP Les instances ASAP [70] sontégalementsontégalement présentées dans un format XML très proche de celui de NRP10. La différence fondamentale réside dans l'absence de r` egle spécifique dans les contrats. Toutes les r` egles sont représentées sous forme de motifs dont le nombre d'apparitions est borné. Ainsi pour représenter la r` egle limitant le nombre de repos pour un employé ,
un motif sera représenté par un automate, et si son nombre d'apparitions est limité, par un automate pondéré. Or, lorsqu'il s'agit de compter simplement une activité, l'emploi d'un compteur simple est plus efficace que l'agrégation d'un automate pondérépondéréà un seuí etat. De plus l'automate représentant les r` egles sur les fins de semaines peut se construire de diversesmanì eres. La représentation de ces r` egles par un motif dans les instances ASAP produit un automate non optimisé (n'intégrant pas de compteur) Afin de réduire l'impact de cette modélisation, nous avons ajoutéajoutéà notre parseur un filtre permettant d'identifier les motifsétantmotifsétant en réalité des compteurs simples ainsi qu'un ensemble de filtres permettant de reconna??trereconna??tre des r` egles génériques exprimées sous forme de motifs. Enfin sur certaines instances, de nouvelles r` egles sont apparues comme la limitation de la consommation de ressources temporelles. Nous ne sommes pas aujourd'hui capables de parser ces r` egles. Cependant, la modularité de notre framework permettrait de les ajouter relativement rapidement. Nous nous sommes tout d'abord attaqués aux instances non sur-contraintes de la librairie. Afin de les résoudre nous avons branché sur la variable représentant la violation duprobì eme. Ainsi, lorsqu'elle est fixéè a zéro, la back-propagation des contraintes multicost-regular permet de rendre « dures, En effet, nous considérons pour les instances NRP10 qu Malgré ces différences de structures, le cadre de résolution basé sur la contrainte multicost-regular proposé au chapitre 9 a permis de trouver d'excellentes solutionsàsolutionsà toutes ces instances ,
Le cadre présenté est en effet capable d'? etre autonome depuis la lecture du fichier d'instance jusqu'` a l'affichage d'une solution Ces expérimentations mettent par ailleurs en avant l'expressivité des automates et la capacité des meta-contraintes-automatesàautomatesà modéliser des r` egles nombreuses et variées Il est notamment plus facile de modifier un automate ou de le créer demanì ere systématique que de multiplier les algorithmes de filtrage pour les diverses contraintes. Les résultats obtenus lorsque la variable de violation est fixée montre que la contrainte multicost-regular, grâcè a l'agrégation d'un ensemble de r` egles intrinsèquement liées, permet un filtrage efficace. Enfin, la qualité des solutions obtenues sur lesprobì emes pures d'optimisation montre l'intérêt de la construction d'heuristiques s'appuyant sur la structure des contraintes. L'agrégation de nombreuses r` egles au sein d'une seule contrainte multicost-regular facilite par ailleurs ce travail. 124 11. Conclusion chapitre 10, sont très encourageants dans la mesure o` u nous nous comparonsàcomparonsà des méthodes développées spécifiquement pour chaque instance, l` a o` u nous utilisons un cadre de modélisation et de résolution générique et automatisé. De plus, nous nous sommes attaqués, notamment sur les instances NRP10, ` a desprobì emes d'optimisation purs, o` u la densité de solutions est très importante. La puissance des contraintes multicost-regular et soft-multicost-regular réside, audeì a de l'aspect modélisation, dans une capacité de filtrage accrue, par rapportàrapportà un modèle décomposé. Or, dans le cadre deprobì emes d'optimisation purs, il est fait usage, Nous obtenons cependant sur ces instances des résultats proches de l'optimal ce qui tendàtendà prouver que l'heuristique s'appuyant sur la structure de multicost-regular propose des solutions de qualité ,
Efficient string matching: an aid to bibliographic search, Communications of the ACM, vol.18, issue.6, p.340, 1975. ,
DOI : 10.1145/360825.360855
A Multiple Objective Nurse Scheduling Model, A I I E Transactions, vol.24, issue.5, pp.55-60, 1981. ,
DOI : 10.1080/05695558108974536
A flexible model and a hybrid exact method for integrated employee timetabling and production scheduling. Practice and Theory of Automated Timetabling VI, pp.67-84, 2007. ,
URL : https://hal.archives-ouvertes.fr/hal-00196330
Mixed Integer Programming to Schedule a Single-Shift Workforce under Annualized Hours, Annals of Operations Research, vol.128, issue.1-4, pp.199-215, 2004. ,
DOI : 10.1023/B:ANOR.0000019105.54898.a4
Integrated days off and shift personnel scheduling, Computers & Industrial Engineering, vol.9, issue.4, pp.395-404, 1985. ,
DOI : 10.1016/0360-8352(85)90027-0
Personnel scheduling with flexshift models, Journal of Operations Management, vol.5, issue.3, pp.327-338, 1985. ,
DOI : 10.1016/0272-6963(85)90017-8
Preference scheduling for nurses using column generation, European Journal of Operational Research, vol.164, issue.2, pp.510-534, 2005. ,
DOI : 10.1016/j.ejor.2003.06.046
Scheduling staff using mixed integer programming, European Journal of Operational Research, vol.98, issue.3, pp.473-484, 1997. ,
DOI : 10.1016/S0377-2217(97)00055-6
Revisiting the cardinality operator and introducing the cardinalitypath constraint family, Logic Programming, pp.59-73, 2001. ,
On matrices, automata, and double counting. Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems, pp.10-24, 2010. ,
URL : https://hal.archives-ouvertes.fr/hal-00915717
Global Constraint Catalog, 2008. ,
URL : https://hal.archives-ouvertes.fr/hal-00485396
Introducing global constraints in CHIP, Mathematical and Computer Modelling, vol.20, issue.12, pp.97-123, 1994. ,
DOI : 10.1016/0895-7177(94)90127-9
URL : https://hal.archives-ouvertes.fr/hal-00442810
Reformulation of Global Constraints Based on Constraints Checkers, Constraints, vol.11, issue.2, pp.339-362, 2005. ,
DOI : 10.1007/s10601-005-2809-x
Combining Tree Partitioning, Precedence, and Incomparability Constraints, Constraints, vol.21, issue.1, pp.459-489, 2008. ,
DOI : 10.1007/s10601-007-9040-x
URL : https://hal.archives-ouvertes.fr/hal-00481533
Théorie des graphes et ses applications, 1958. ,
A multi-objective approach to nurse scheduling with both hard and soft constraints, Socio-Economic Planning Sciences, vol.30, issue.3, pp.183-193, 1996. ,
DOI : 10.1016/0038-0121(96)00010-9
Among, common and disjoint constraints. Recent Advances in Constraints, pp.29-43, 2006. ,
URL : https://hal.archives-ouvertes.fr/lirmm-00378928
MAC and combined heuristics: Two reasons to forsake FC (and CBJ?) on hard problems, Principles and Practice of Constraint Programming?CP96, pp.61-75 ,
DOI : 10.1007/3-540-61551-2_66
Constraint propagation, 2006. ,
Reformulating global constraints : The SLIDE and REGULAR constraints, SARA, pp.80-92, 2007. ,
A goal programming approach to strategic resource allocation in acute care hospitals, European Journal of Operational Research, vol.140, issue.3, pp.541-561, 2002. ,
DOI : 10.1016/S0377-2217(01)00219-3
hibiscus: A Constraint Programming Application to Staff Scheduling in Health Care, Principles and Practice of Constraint Programming, pp.153-167, 2003. ,
DOI : 10.1007/978-3-540-45193-8_11
Subgradient methods. lecture notes of EE392o, 2003. ,
Continuous personnel scheduling algorithms : a literature review, Journal of the Society for Health Systems, vol.2, issue.2, p.8, 1991. ,
Encodings of the Sequence Constraint, Principles and Practice of Constraint Programming ? CP 2007, pp.210-224, 2007. ,
DOI : 10.1007/978-3-540-74970-7_17
New methods to color the vertices of a graph, Communications of the ACM, vol.22, issue.4, pp.251-256, 1979. ,
DOI : 10.1145/359094.359101
A memetic approach to the nurse rostering problem, Applied Intelligence, vol.15, issue.3, pp.199-214, 2001. ,
DOI : 10.1023/A:1011291030731
Fitness evaluation for nurse scheduling problems, Proceedings of the 2001 Congress on Evolutionary Computation (IEEE Cat. No.01TH8546), pp.1139-1146, 2001. ,
DOI : 10.1109/CEC.2001.934319
URL : http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.2.7890
A Tabu-Search Hyperheuristic for Timetabling and Rostering, Journal of Heuristics, vol.9, issue.6, pp.451-470, 2003. ,
DOI : 10.1023/B:HEUR.0000012446.94732.b6
Manpower Scheduling With Variable Demands And Alternate Weekends Off, INFOR: Information Systems and Operational Research, vol.16, issue.2, pp.101-111, 1978. ,
DOI : 10.1080/03155986.1978.11731694
A genetic algorithm for scheduling staff of mixed skills under multi-criteria, European Journal of Operational Research, vol.125, issue.2, pp.359-369, 2000. ,
DOI : 10.1016/S0377-2217(99)00391-4
The choco constraint solver, INFORMS Annual meeting, 2006. ,
Modeling and Solving the Crew Rostering Problem, Operations Research, vol.46, issue.6, pp.820-830, 1998. ,
DOI : 10.1287/opre.46.6.820
Nurse rostering problems??????a bibliographic survey, European Journal of Operational Research, vol.151, issue.3, pp.447-460, 2003. ,
DOI : 10.1016/S0377-2217(03)00021-3
A constraint-based nurse rostering system using a redundant modeling approach, ictai, p.140, 1996. ,
Three models for the description of language. Information Theory, IRE Transactions on, vol.2, issue.3, pp.113-124, 1956. ,
Syntactic structures, 2002. ,
DOI : 10.1515/9783110218329
Nurse rostering at the hospital authority of Hong Kong, PROCEEDINGS OF THE NATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE, pp.951-956, 1999. ,
Modeling the regular constraint with integer programming, Proceedings of the 4th international conference on Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems, CPAIOR '07, pp.29-43, 2007. ,
Formal languages for integer programming modeling of shift scheduling problems, Constraints, pp.1-23, 2007. ,
Horoplan : computer-assisted nurse scheduling using constraint-based programming, Journal of the Society for Health Systems, vol.5, issue.1, pp.41-54, 1995. ,
A Cost-Regular Based Hybrid Column Generation Approach, Constraints, vol.12, issue.4???5, pp.315-333, 2006. ,
DOI : 10.1007/s10601-006-9003-7
URL : https://hal.archives-ouvertes.fr/hal-00293562
Solving a nurse scheduling problem with knapsacks, networks and tabu search, Journal of the Operational Research Society, vol.51, issue.7, pp.825-833, 2000. ,
DOI : 10.1057/palgrave.jors.2600970
Handbook of Weighted Automata, 2009. ,
DOI : 10.1007/978-3-642-01492-5
A distributed genetic algorithm for employee staffing and scheduling problems, Proceedings of the 5th International Conference on Genetic Algorithms, p.367 ,
An Annotated Bibliography of Personnel Scheduling and Rostering, Annals of Operations Research, vol.127, issue.1-4, pp.21-144, 2004. ,
DOI : 10.1023/B:ANOR.0000019087.46656.e2
Personnel scheduling in laboratories. Practice and Theory of Automated- Timetabling IV, pp.113-119, 2003. ,
Backtrack Programming, Journal of the ACM, vol.12, issue.4, pp.516-524, 1965. ,
DOI : 10.1145/321296.321300
A dual algorithm for the constrained shortest path problem, Networks, vol.17, issue.4, pp.293-310, 1980. ,
DOI : 10.1002/net.3230100403
Increasing tree search efficiency for constraint satisfaction problems, Artificial Intelligence, vol.14, issue.3, pp.263-313, 1980. ,
DOI : 10.1016/0004-3702(80)90051-X
Introduction to Automata Theory, Languages, and Computation. Pearson Education, 2001. ,
Evolutionary algorithms for nurse scheduling problem, Proceedings of the 2000 Congress on Evolutionary Computation. CEC00 (Cat. No.00TH8512), pp.196-203, 2000. ,
DOI : 10.1109/CEC.2000.870295
Unification and extension of weighted finite automata applicable to image compression, Theoretical Computer Science, vol.302, issue.1-3, pp.275-294, 2003. ,
DOI : 10.1016/S0304-3975(02)00846-0
Efficient context-free grammar constraints, AAAI, pp.310-316, 2008. ,
Reducibility among combinatorial problems. 50 Years of Integer Programming, pp.219-241, 1958. ,
Representation of Events in Nerve Nets and Finite Automata, Automata studies, vol.34, pp.3-41, 1956. ,
DOI : 10.1515/9781400882618-002
Nurse scheduling on a microcomputer, Computers & Operations Research, vol.18, issue.8, pp.731-739, 1991. ,
DOI : 10.1016/0305-0548(91)90011-F
A general multi-shift scheduling system, Journal of the Operational Research Society, vol.35, issue.11, pp.1208-1217, 2004. ,
DOI : 10.1080/002075497195010
Flow-Based Propagators for the SEQUENCE and Related Global Constraints, Proceedings of CP'2008, pp.159-174, 2008. ,
DOI : 10.1007/978-3-540-85958-1_11
Cyclical scheduling and allocation of nursing staff, Socio-Economic Planning Sciences, vol.7, issue.5, pp.471-487, 1973. ,
DOI : 10.1016/0038-0121(73)90043-8
Sequencing and Counting with the multicost-regular Constraint, 6th International Conference on Integration of AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems (CPAIOR'09), pp.178-192, 2009. ,
DOI : 10.1007/978-3-642-01929-6_14
URL : https://hal.archives-ouvertes.fr/hal-00394434
Solving nurse rostering problems using soft global constraints. Principles and Practice of Constraint Programming -CP, pp.73-87, 2009. ,
Solving rostering tasks as constraint optimization. Lecture notes in computer science, pp.191-212, 2001. ,
Cyclic and non-cyclic scheduling of 12 h shift nurses by network programming, European Journal of Operational Research, vol.104, issue.3, pp.582-592, 1998. ,
DOI : 10.1016/S0377-2217(97)00006-4
Speech Recognition with Weighted Finite-State Transducers, Computer Speech & Language, vol.16, issue.1, pp.69-88, 2002. ,
DOI : 10.1007/978-3-540-49127-9_28
Goal Programming Model Subsystem of a Flexible Nurse Scheduling Support System, IIE Transactions, vol.5, issue.3, pp.306-316, 1988. ,
DOI : 10.1287/opre.2.3.339
A Filtering Algorithm for the Stretch Constraint, Principles and Practice of Constraint Programming?CP 2001, pp.183-195, 2001. ,
DOI : 10.1007/3-540-45578-7_13
Constraint-based rostering, The 7th International Conference on the Practice and Theory of Automated Timetabling, PATAT, 2008. ,
A Regular Language Membership Constraint for Finite Sequences of Variables, Proceedings of CP'2004, pp.482-495, 2004. ,
DOI : 10.1007/978-3-540-30201-8_36
A large neighbourhood search approach to??the?? multi-activity shift scheduling problem, Journal of Heuristics, vol.10, issue.2, pp.373-392, 2010. ,
DOI : 10.1007/s10732-009-9106-6
Decomposing global grammar constraints. Principles and Practice of Constraint Programming ?CP, pp.590-604, 2007. ,
DOI : 10.1007/978-3-540-74970-7_42
URL : http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.87.1320
Impact-based search strategies for constraint programming. Principles and Practice of Constraint Programming?CP, pp.557-571, 2004. ,
The cardinality matrix constraint. Principles and Practice of Constraint Programming?CP, pp.572-587, 2004. ,
A filtering algorithm for global sequencing constraints. Principles and Practice of Constraint Programming-CP97, pp.32-46, 1997. ,
A filtering algorithm for constraints of difference in csps, AAAI '94 : Proceedings of the twelfth national conference on Artificial intelligence, pp.362-367, 1994. ,
Generalized arc consistency for global cardinality constraint, Proceedings of the 13th National Conference on AI (AAAI/IAAI'96), pp.209-215, 1996. ,
Automata for nogood recording in constraint satisfaction problems. Integration of SAT and CP Techniques, p.113, 2006. ,
Inside case-based reasoning, 1989. ,
Handbook of constraint programming, 2006. ,
Between finite state and prolog : constraint-based automata for efficient recognition of phrases, Nat. Lang. Eng, vol.2, pp.365-366, 1996. ,
The theory of grammar constraints. Principles and Practice of Constraint Programming-CP, pp.530-544, 2006. ,
Theoretical foundations of CP-based lagrangian relaxation. Principles and Practice of Constraint Programming ?CP, pp.634-647, 2004. ,
Ruszcay`nski. Minimization methods for non-differentiable functions, 1985. ,
A fuzzy set theory approach to the aircrew rostering problem. Fuzzy sets and systems, pp.261-271, 1998. ,
On Manpower Scheduling Algorithms, SIAM Review, vol.24, issue.3, pp.275-287, 1982. ,
DOI : 10.1137/1024063
On computable numbers, with an application to the Entscheidungsproblem, Proceedings of the London Mathematical Society, p.230, 1937. ,
Workforce planning in synchronous production systems, European Journal of Operational Research, vol.136, issue.3, pp.551-572, 2002. ,
DOI : 10.1016/S0377-2217(01)00056-X
Revisiting the Sequence Constraint, Proceedings of CP'2006, pp.620-634, 2006. ,
DOI : 10.1007/11889205_44
On global warming: Flow-based soft global constraints, Journal of Heuristics, vol.1, issue.4, pp.347-373, 2006. ,
DOI : 10.1007/s10732-006-6550-4
An advanced model and novel meta-heuristic solution methods to personnel scheduling in healthcare, 2002. ,
On the characterization and generation of nurse scheduling problem instances, European Journal of Operational Research, vol.196, issue.2, pp.457-467, 2009. ,
DOI : 10.1016/j.ejor.2008.03.044
Scheduling Nursing Personnel According to Nursing Preference: A Mathematical Programming Approach, Operations Research, vol.24, issue.5, pp.842-856, 1976. ,
DOI : 10.1287/opre.24.5.842
A Mathematical Programming Model for Scheduling Nursing Personnel in a Hospital, Management Science, vol.19, issue.4-part-1, pp.411-422, 1972. ,
DOI : 10.1287/mnsc.19.4.411
Constraint programming for nurse scheduling, IEEE Engineering in Medicine and Biology Magazine, vol.14, issue.4, pp.417-422, 1995. ,
DOI : 10.1109/51.395324
Scheduling doctors for clinical training unit rounds using tabu optimization. Practice and Theory of AutomatedTimetabling IV, pp.120-128, 2003. ,
STAFFING THE NURSING UNIT, Nursing Research, vol.14, issue.3, p.236, 1965. ,
DOI : 10.1097/00006199-196501430-00011
STAFFING THE NURSING UNIT Part II. The Multiple Assignment Technique, Nursing Research, vol.14, issue.4, p.299, 1965. ,
DOI : 10.1097/00006199-196501440-00004