Parallel optimisation in the SCOOP library
This paper shows how parallelism has been integrated into SCOOP, a C++ class library for solving optimisation problems. After a description of the modeling and the optimisation parts of SCOOP, two new classes that permit parallel optimisation are presented: a class whose only purpose is to handle messages and a class for managing optimiser and message handler objects. Two of the most interesting aspects of SCOOP, modularity and generality, are preserved by clearly separating problem representation, solution techniques and parallelisation scheme. This allows the user to easily model a problem and construct a parallel optimiser for solving it by combining existing SCOOP classes.
Unable to display preview. Download preview PDF.
- 1.MPI Forum. MPI: A message-passing interface standard. International Journal of Supercomputer Applications, 8 (3/4):165–416, 1994.Google Scholar
- 2.SCOOP 2.0 Reference Manual, SINTEF report no. STF 42A98001, ISBN 82-1400047-5, 1998.Google Scholar
- 3.M. Benaichouche, V. D. Cung, S. Dowaji, B. Le Cun, T. Mautor and C. Roucairol. Building a Parallel Branch and Bound Library. Solving Combinatorial Optimization Problems in Parallel, LNCS 1054, Springer, 201–231, 1996.Google Scholar
- 5.J.H. Holland. Adaptation in Natural and Artificial Systems. The University of Michigan Press, 1975.Google Scholar
- 6.S. Kirkpatrick, C. Gellat and M. Vecchi. Optimisation by simulated annealing. Science, 220:671–680, 1983.Google Scholar
- 7.G. Misund, G. Hasle and B. S. Johansen. Solving the clear-cut scheduling problem with geographical information technology and constraint reasoning. In ScanGIS'95 Proceedings, pages 42–56, 1995. Also published as SINTEF report no. STF 33S95027.Google Scholar