智能混合系统  

A Review of Intelligent Hybrid Systems

作　　者：刘振凯[1] 贵忠华[2] 蔡青[2] 

机构地区：[1]西安交通大学 [2]西北工业大学

出　　处：《机电一体化》1998年第6期15-18,共4页Mechatronics

摘　　要：作为人工智能的两个主要研究方向,专家系统和神经网络在过去的几十年中均取得了很大的成功,但二者又存在一些各自的缺陷,不能单独解决一些复杂的工程实际问题。因此,智能混合系统就成为当今人工智能研究领域中的一个非常重要且具有很大实用价值和广阔发展前景的研究领域。本文分析比较了专家系统和神经网络的基本特点,论述智能混合系统的研究现状并指出存在的一些问题。Expert system and neural network are two principal developing fields of artificial intelligence. They have both made great progress and successes in the last decades. However each of them has its own drawbacks and can not deal with complex engineering problems. So, it becomes a very important research field with great practical value and broad developing prospects to research intelligent hybrid system based on neural networks and expert systems. In this paper,the characteristics of expert systems and neural networks are analyzed, a survey of intelligent hybrid systems is made and some problems are pointed out.

关 键 词：人工智能 专家系统 神经网络 混合系统 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]