摘要
基于认知科学的观点 ,提出了一类智能控制的认知框架 ,它由智能计算结构和人构成 ,能实现人机结合、智能互补。其中的智能计算结构是一种包含两级处理的紧耦合结构 ,前级通过非线性变换 ,实现对输入信息的扩展与增强 ;后级完成中枢处理的动态局域网络 ,能实现信息空间的优化分解 ,降低计算复杂性 ,提高效率和实时性。该文分析了框架的功能及有关环节的实现手段 。
From the viewpoint of recognition science, a recognitive frame, which is composed of a computing configuration with intelligence and human, is proposed for intelligent control in this paper. In this frame human machine combination and intelligence complementarity can be implemented. The computing configuration with intelligence is a compact coupling structure with a two level processing unit. In the front level input information is expanded and enhanced by a nonlinear transform. While in the behind level, a dynamic network is used as a central processing unit, and information space can be optimally decomposed so as to reduce computing complexity, and improve efficiency and real time. Subsequently, some functions and implementation methods are discussed for the recognitive frame. Also a designed example is shown in this paper.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2001年第1期48-51,共4页
Proceedings of the CSEE
基金
国家重点基础研究专项经费资助项目! (G19980 30 40 5 )
