【24h】

Autonomously creating a hierarchy of intelligent agents using clustering in a multi-agent system

机译:在多主体系统中使用集群自主创建智能主体的层次结构

获取原文

摘要

In this paper, we present an approach to automatically create a hierarchy of agents. The hierarchy clusters similar agents and, thereby, optimizes locality. The clustering improves task efficiency in the multi-agent system. Moreover, clustering reduces communication costs between closely connected agents and speeds up searching for data in networks and graphs. In addition, the approach supports task handling by the agents and reasoning at different levels in hierarchy by dividing the overall task among the agents in the hierarchy. The benefit of dividing tasks into small separate pieces is that the agents can work with the task independently using parallel computation. The results from the computation can be efficiently reassembled by the hierarchy of agents.
机译:在本文中,我们提出了一种自动创建代理层次结构的方法。层次结构将相似的代理程序聚集在一起,从而优化位置。集群提高了多代理系统中的任务效率。此外,群集降低了紧密联系的代理之间的通信成本,并加快了在网络和图形中搜索数据的速度。另外,该方法通过将整个任务划分到层次结构中的各个代理之间,从而支持了代理的任务处理和层次结构中不同级别的推理。将任务分成小块的好处是,代理可以使用并行计算独立地处理任务。计算的结果可以通过代理层次结构有效地重新组合。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号