page 1  (18 pages)
2to next section

Constraint Logic Programming on

Boolean, Integer and Real Intervals

Fr?ed?eric Benhamou?, William J. Older and Andr?e Vellinoy

Accepted for publication in Journal of Symbolic Computation

September 21, 1994

Abstract

Imperative programming languages for computing on intervals stand in contrast to systems of relational interval arithmetic that are seamlessly integrated into a logic programming language like CLP(BNR). The combined power of a symbolic, logic programming language on the one hand and a mathematically correct, logically sound and computationally efficient programming system (constraint arithmetic on intervals) on the other, is considerably more powerful than either taken separately. Boolean, integer and real-valued simultaneous constraint equations can be mixed freely in CLP(BNR), not only to contain floating point errors and perform sensitivity analysis but also to express a wide variety of linear and non-linear programming problems, scheduling and configuration problems as well as optimization and operations research problems. Some simple examples of how this can be done are given as well as a summary report on experimental applications of this technology to industrial problems.

?Groupe d'Intelligence Artificielle, Facult?e des Sciences de Luminy, case 901,163, avenue de Luminy, 13288 Marseille Cedex 9 FRANCE, benham@univ-orleans
yBell Northern Research, Computing Research Laboratory, PO Box 3511, Station C, K1Y 4H7 Ottawa, Ontario, Canada, fvellino,wolderg@bnr.ca