实时控制神经网络专家系统结构及推理算法

来源 :模式识别与人工智能 | 被引量 : 0次 | 上传用户:yong5665
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
为了弥补基于单一人工智能方法所形成的智能控制系统的不足,提高基于产生式规则的控制专家系统的性能,本文提出了一种基于神经网络推理机及结合型外部知识表示的神经网络专家系统模型、基本结构及工作原理;在此基础上提出了一种实时控制神经网络专家系统(RTNNES)结构及其知识表示技术;提出了联合使用神经网络与专家系统技术的实时正向推理算法,基于事件的非单调推理算法。 In order to make up for the shortage of intelligent control system based on a single artificial intelligence method and improve the performance of control expert system based on production rules, this paper proposes a model of neural network expert system based on neural network reasoner and combined external knowledge representation , The basic structure and working principle of RTNNES are presented. Based on this, a real-time control neural network expert system (RTNNES) structure and its knowledge representation technology are proposed. A real-time forward reasoning algorithm based on joint use of neural network and expert system is proposed. Non-monotonic reasoning algorithm.
其他文献
会议
期刊
期刊
期刊
期刊