作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2006, Vol. 32 ›› Issue (21): 233-234,. doi: 10.3969/j.issn.1000-3428.2006.21.081

• 工程应用技术与实现 • 上一篇    下一篇

一种基于多Agent的智能家居系统研究与设计

李晓华1,赵现纲2   

  1. (1. 北京农学院计算机教研室,北京 102206;2. 北京邮电大学计算机学院,北京 100876)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2006-11-05 发布日期:2006-11-05

Research and Design on MA-based Intelligent Home System

LI Xiaohua1, ZHAO Xiangang2   

  1. (1. Department of Computer Science, Beijing Agriculture College, Beijing 102206; 2. School of Computer, Beijing University of Posts and Telecommunications, Beijing 100876)
  • Received:1900-01-01 Revised:1900-01-01 Online:2006-11-05 Published:2006-11-05

摘要: 就Agent技术在智能控制领域的应用做了一次尝试,设计出了一种实用性强、较综合全面的基于多Agent的智能家居MAIHS系统方案。考虑到家庭环境在计算、存储以及通信方面的局限性,该系统被设计为2层,由2种Agent组成:监督者(SA)和协作者(CA),并采用集中与分布式相结合的协调策略。此外,系统还将传统的账号密码机制与基于人体生物特征的身份认证方式相结合,使用3层BP神经网络进行综合身份验证,从而可确保拥有较高的保密性、可认证性和灵活性等性能。

关键词: 多智能体, 智能家居, 指数平滑模型, 生物特征识别, J2ME

Abstract: This paper presents a new attempt to apply agent technology into intelligent control domain and proposes a practicable, integrative prototype of multi agent based intelligent home system(MAIHS). Taking into account the limits of household circumstance in computing, storing and communicating, this system is designed on two-layered architecture and works with two agent types: surveillant agents(SA) and cooperating agents(CA), and they communicate under the integration of centralized-distributed strategy. Furthermore, with the general account-password recognition mechanism together with biometrics fulfilling the first-level certification, a three-leveled BP neural network makes the final decision that the system can enjoy high secrecy, authentication and flexibility.

Key words: Multi agent, Intelligent home, Exponential smoothing model, Biometrics, J2ME

中图分类号: