Extraction of Implicit Resource Relationships in Multi-agent Systems

摘要: This paper discusses the storage and analysis of past hierarchical-planning results in order to identify implicit costs resource relationships between activities multi-agent contexts. We have previously proposed a plan-reuse framework which plans are stored as templates after use then reused speed up planning activity systems. In this paper, we propose mechanizm for learning, from that consist used data recorded during execution, concerning usage by multiple agents. Here, indicates exist environments where agents deployed but not described domain models have. The also provides guidance on planner executor should record when learned rules be applied. Finally, some examples show how learning enables creation more appropriate solutions