摘要
提出了一种基于聚类算法和模糊神经网络的非线性模拟电路故障诊断方法。通过一个无监督的聚类算法自组织地确定模糊规则的数目并生成一个初始的故障诊断模糊规则库,构造了一类模糊神经网络,通过训练调整网络权值,使故障诊断模糊规则库的分类更加精确,并通过仿真实验验证了该方法的有效性。
An approach for the fault diagnosis of nonlinear analog circuits based on clustering and fuzzy neural network is presented. Through an unsupervised clustering technique, the number of fuzzy rules is determined and an initial fuzzy rule is generated from the given input-output data. A kind of fuzzy neural networks is constructed and its weights are tuned to make the parame- ters of the constructed fuzzy rule of the fault diagnosis base more precise. The availability of the method is examined by simulated tests.
出处
《微电子学与计算机》
CSCD
北大核心
2006年第8期1-3,共3页
Microelectronics & Computer
基金
国家自然科学基金项目(50277010)
湖南省自然科学基金项目(04JJ6034)
高等院校博士学科点专项科研基金项目(20020532016)
湖南省科技计划项目(04FJ2003
03GKY3115)
关键词
故障诊断
神经网络
模糊规则
参数聚类
非线性电路
Fault diagnosis, Neural network, Fuzzy rule, Parameter cluster, Nonlinear circuits
