首页 | 官方网站   微博 | 高级检索  
     

故障诊断交流字典法的前向神经网络实现方法
引用本文:崔莼,罗先觉.故障诊断交流字典法的前向神经网络实现方法[J].微电子学,1996,26(5):313-318.
作者姓名:崔莼  罗先觉
作者单位:西安交通大学
基金项目:高校博士学科点科研基金
摘    要:提出了一种采用BP算法的前向多层神经网络实现模拟电路故障诊断交流字典的方法。

关 键 词:模拟电路  故障诊断  神经网络

Implementation of AC Fault Dictionary by Feed Forward Multilayer Neural Network
Abstract:A method to construct an AC fault dictionary for diagnosis of analog circuits is presented,which is based on the feed forward multilayer neural network (FNN)using backpropagation algorithm(BP).The basic conception is to implement the fault dictionary within the FNN.In the proposed technique,only a small number of samples is needed to train the FNN,and the effects on correct fault location due to component tolerances can be minimized. The computational time after test is likely to be significantly reduced by the use of FNN and massive parallel processing of the diagnosis.
Keywords:Analog  circuit  Neural  network  Fault  diagnosis  
本文献已被 CNKI 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号