专家神经网络融合的集成控制系统研究及应用  

Research and Application on Integration Control System Based on Expert System and Neural Network Fusion

作　　者：杜明芳[1] 范同顺[1] 李红星[1] 

机构地区：[1]北京联合大学自动化学院,北京100101

出　　处：《电气传动》2009年第8期36-39,共4页Electric Drive

基　　金：北京市教育委员会科技计划面上项目(KM200811417012)

摘　　要：综合应用神经网络和专家系统的相关理论对复杂工业控制系统集成问题作了探讨。给出了专家神经网络混合集成控制系统的总体架构。借助离散神经网络概念提出了集成系统中的特定神经元表示方法,给出了特定神经网络表达形式。介绍了集成系统专家知识库的构建思路,并以一个实例详细说明了行业集成专家知识库的设计方法。实际设计构建过程表明专家神经网络融合方法适合于描述复杂集成化控制系统。The complex integration control problem was discussed together using relative neural network and expert system theory. The expert system and neural network hybrid framework of integration system was given. The specific description method of neurons based on discrete neural network definitions was put forward. And the expression of the neuron was given. The design thinking of expert system knowledge base of integration system was introduced. And the design method of expert system knowledge base was shown by an example. The factual design shows the expert system and neural system fusion method is suitable to describe complex integration control system.

关 键 词：离散神经网络 专家系统 融合 集成控制 知识库 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]