摘要
由于模拟电路存在容差性、元件参数连续可变性和非线性等因素,且实际中也受到可测试点数量的限制,基于传统模拟电路故障诊断法在实际工程中难以取得理想的效果。而神经网络具有容错性、泛化能力和非线性处理能力等特点,本文针对雷达电路的故障进行快速有效的特征提取,构造神经网络样本,并结合电路的频率特性来解决模拟电路故障诊断中存在的问题。
Analog circuit has its tolerance on component parameters, continuous response and nonlinearity, and the limitation of the number of test points, it is difficult to achieve expected results in practical engineering based on classical circuit fault diagno- sis. While neural network has the characteristics of fault tolerance, generalization ability and non-linear processing, this paper ex- tracts the fault feature in radar circuit rapidly and validly, constructs the samples of neural network, and solves the problems in fault diagnosis of analog circuits with frequency characteristics of circuits.
出处
《计算机与现代化》
2012年第6期27-30,共4页
Computer and Modernization
关键词
模拟电路
故障诊断
特征提取
神经网络
频率特性
analog circuit
fault diagnosis
feature extraction
nelrral network
frequency characteristics
