摘要
提出了一种采用改进的 BP快速算法实现模拟电路软故障诊断的方法 .文中对构造神经网络交流故障字典的过程给出了详细说明 ,特别是隐层节点数确定 ,测试信号频率的优选及 MATL AB神经网络工具箱的使用 .实例表明 。
A modified BP fast algorithm for fault diagnosis of analog circuits is introduced in this paper.The establishment of the ANN alternating fault dictionary for analog circuits is exhaustively discussed,especially the determination of the node number of hidden layer,the selection of testing signal frequency,and the use of toolbox with MATLAB 5.1.An example showing this novel diagnosing approach to analog circuit soft fault is given.
出处
《内蒙古工业大学学报（自然科学版）》
2000年第2期130-133,共4页
Journal of Inner Mongolia University of Technology：Natural Science Edition
关键词
模拟电路
故障诊断
人工神经网络
BP算法
analog circuit
fault diagnosis
ANN
BP algorithm
AC fault dictionary
