摘要
给出了离散Hopfield神经网络结构和模型．利用Hopfield神经网络的演变过程是一种计算联想记忆的过程,它适用于正交(或近似正交)模式的记忆性质,给出了一种Hopfield神经网络的双向联想记忆模式的记忆矩阵构造方法,并提出了一种改进的基于Hopfield神经网络的控制系统故障诊断的算法,利用此算法实现实时检测混烧控制系统的故障和异常,对混烧控制系统的调节器和阀门进行故障诊断和故障信息提示．
This paper presents a model and architecture of discrete Hopfield neural networks, By using the characters of discrete Hop field neural networks, whose evolvement procedure is a computing course of associative memory suitable to memorize the orthogonal model, this paper proposes a method for constructing memory matrix based on bi-directional associative memory of Hopfield neural networks. It also discusses an improved failure detection algorithm based on Hopfield neural networks for control system. This algorithm has been applied to realize the fault and abnormity detection for the mixture burning control system on line, and has given the fault information of the adjustors and the values in this control system.
出处
《湖南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2007年第3期33-35,共3页
Journal of Hunan University:Natural Sciences
基金
国家自然科学基金(50277010
60673084)
教育部高等学校博士学科点专项科研基金(20020532016)
湖南省科技计划项目(04FJ2003
03GKY3115
05GK2005)
关键词
故障诊断
神经网络
混烧系统
failure diagnosis
neural networks
mixture burning control system
