论文信息 - Impact- and Cost-Oriented Propagator Scheduling for Faster Constraint Propagation

Impact- and Cost-Oriented Propagator Scheduling for Faster Constraint Propagation

Constraint Propagation can be speeded up significantly by choosing a good execution order for propagators. A propagator is an implicit representation of a constraint which is widely used in todays powerful constraint solvers. In this paper we evaluate different ways to find good execution orders automatically during runtime. We extend previous work in this area by two new techniques: fair-scheduling and impact-oriented prioritization of propagators.

Georg Ringwelski | Matthias Hoche

[1] Krzysztof R. Apt,et al. The Essence of Constraint Propagation , 1998, Theor. Comput. Sci..

[2] Eugene C. Freuder,et al. Ordering Heuristics for Arc Consistency Algorithms , 1992 .

[3] François Fages,et al. Experiments in Reactive Constraint Logic Programming , 1998, J. Log. Program..

[4] Philippe Codognet,et al. Compiling Constraints in clp(FD) , 1996, J. Log. Program..

[5] J. Christopher Beck,et al. Trying Again to Fail-First , 2004, CSCLP.

[6] Peter J. Stuckey,et al. Speeding Up Constraint Propagation , 2004, CP.

[7] Georg Ringwelski. Asynchrones Constraintlösen: ein generisches Ausführungsmodell zur adaptiven, inkrementellen Constraintverarbeitung , 2003 .

[8] Thom W. Frühwirth,et al. Constraint Handling Rules , 2009, Constraint Programming.

[9] Patrick Cousot,et al. Automatic synthesis of optimal invariant assertions: Mathematical foundations , 1977 .

[10] Arnaud Lallouet,et al. Intermediate Consistencies by Delaying Expensive Propagators , 2004, FLAIRS Conference.