摘要： 根据决策支持系统应用中存在的问题，提出了一种新的利用决策支持系统求解问题的思路，并将Agent思想和技术引入到这一思路的实现中，设计了一种基于Agent的智能决策支持系统的架构，该架构主要由模型选择Agent、模型构造Agent、模型求解Agent和人机交互Agent 4个部分组成。该结构的实现可有效地突破基于模型的决策支持系统在求解问题时难以适应动态环境变化的障碍，使决策支持系统真正成为面向问题系统，极大地提高了决策支持系统的智能型，能够根据问题的变化作出适应性的自主的调整，满足用户的求解问题的要求。

关键词: 智能决策支持系统, Agent, 模型, 人机交互

Abstract: This paper puts forward a new approach to solve the problem in decision support system which overcomes the disadvantage of current system. The paper designs the structure of intelligent decision support system by connecting the agent theory and practice with this approach. It consists of model selecting agent, model-constructing agent, model solving agent, and human computer interactive agent. This overcomes the difficulty in adapting the environmental changes when solving problem using DSS based on model and turns the DSS to be a problem oriented system and improves the intelligence of the system thus enables the system to make some adjustments according the changes of the problem and meet the requirements of the user.

Key words: Intelligent decision support system, Agent, Model, Human computer interaction