A/sup 2/: an agent-oriented programming architecture for multi-agent constraint satisfaction problems
E.H. Freeman · [1990] Proceedings of the 2nd International IEEE Conference on Tools for Artificial Intelligence · 2002
An agent-oriented programming metaphor is used to extend the analytic capabilities of a constraint logic programming system, such as CLP(R), to the domain of multi-agent constraint satisfaction problems. The resulting implementation provides a set of system primitives, which support at a rudimentary level, the maintenance of private knowledge bases, inter-agent communications, constraint driven multi-agent consensus formation, functional inheritance via 'cloning' and a choice of inheritance lattice search optimization mechanisms, allowing knowledge engineers to make speed vs. flexibility and functional dependence vs. independence trade-offs. A general architecture for agent-oriented programming systems is presented, and some of the more salient aspects of its CLP(R) implementation are summarized.>