Thoughts on Partitioning Large-Scale Configuration Problems

Most ’interesting’ configuration problems tend to be large-scale. Therefore, the configuration networks grow to remarkable sizes and techniques must be considered to keep handling, solving and checking of these networks practically feasible. This paper presents a technique we use in our current configuration project: the partitioning of large problems into pieces of manageable sizes. The very basics of our component-oriented framework for representing configuration domains are presented. Partitioning a configuration problem means identifying subproblems in the configuration graph and solving them separately. Beside the reduction in problem size we provide the user with control mechanisms by displaying a sub-system-graph to guide the user through the configuration process.

[1]  Andreas Günter,et al.  PLAKON—an approach to domain-independent construction , 1989, IEA/AIE '89.

[2]  Sanjay Mittal,et al.  Towards a Generic Model of Configuraton Tasks , 1989, IJCAI.

[3]  Rina Dechter,et al.  Tree Clustering for Constraint Networks , 1989, Artif. Intell..

[4]  M. Heinrich,et al.  A resource-based paradigm for the configuring of technical systems from modular components , 1991, [1991] Proceedings. The Seventh IEEE Conference on Artificial Intelligence Application.

[5]  Eugene C. Freuder,et al.  The Complexity of Some Polynomial Network Consistency Algorithms for Constraint Satisfaction Problems , 1985, Artif. Intell..

[6]  Rina Dechter,et al.  Tree Decomposition with Applications to Constraint Processing , 1990, AAAI.

[7]  A. Haselbock,et al.  COCOS/spl minus/a tool for constraint-based, dynamic configuration , 1994, Proceedings of the Tenth Conference on Artificial Intelligence for Applications.