基于大数据流的Multi-Agent系统模型研究

Study on Multi-Agent System Model Based on Heavy Data Stream

在智能决策支持系统(IDSS)中,由于许多推理模型和方法以巨大的数据量为基础,决策响应速度成为系统的瓶颈。结合软件开发实际,提出了一种基于速度优化的Multi-Agent模型,将系统Agent分为智能交互界面Agent、数据分析处理Agent、数据访问控制Agent三类,通过各Agent的分工与协作来实现系统数据的快速交互;并在ERP采购成本分析子系统中进行建模研究,结果表明该模型能大大缩短系统决策响应时间。

In intelligent decision support system （IDSS）, many reasoning models and methods based on the tremendous amount of data, the response speed of decision- making as a bottleneck. Considered the practice of software development, proposed a M.AS model based on venous optics. The agent was divided into intelligent interface agent, data processing agent and data access control agent. Through agent division and cooperation make the data exchanged rapidly, its application in ERP purehase cost analysis sub- system show that it can make the response time of system decision- making to be shorter.