动态不确定环境下的决策:一种分层决策模型

Making Decision under Real-Time, Unpredicted Environments:a Multi-Layer Model

本文提出一种智能体分层决策结构模型,试图通过分层决策技术有效地解决动态、不确定环境中的智能体的实时决策问题。本模型的高层采用BDI结构,以便为较长期任务的规划和推理提供充分的支持;模型的底层采用反应式结构,以保证对短期实时任务的及时响应。实验结果表明了这种分层模型在某些复杂任务领域中的有效性。

This paper describes our work in developing an agent architecture based on a multi-layer model which will help to solve the problem of real time decision-making in dynamic, unpredicted environments. The high levels of this architecture use BDI model to support long-term planning and reasoning while use reactive model in the lower level which ensures the system to react timely in real-time environments. Experiments show that such a multi-layer archi- tecture is effective in some complicated real-time domains.