人工神经网络在模拟电路故障诊断中的应用

容差和非线性特性使得模拟电路的故障诊断趋于复杂化，针对这两个问题，几年来，人们提出了许多相关的诊断方法。但是，迄今为止．尚没有一种通用的、行之有效的方法。因此，尝试将自组织竞争网络引入到模拟电路的诊断过程中，并提出一种基于竞争网络的诊断方法。由于自组织竞争网络在非线性映射、容错与自学习方面的独特优势，理论上对于解决模拟电路的容差和非线性问题是非常适合的。最后，实例表明，基于竞争网络的方法能够准确识别出电路中的故障。

Application of Artificial Neural Network in Fault Diagnosis of Analog Circuits

The tolerance and nonlinearity in analog circuits make its fault diagnosis difficult. In order to solve this problem, many methods are presented. However, no method can be suitable to all types of analog circuits by far. In this paper, the competitive layer is introduced to the process of fault diagnosis for analog circuits in order to avoid tolerance. A diagnostic method based on competitive layer is presented. At last, an example is presented to iUustrate the method.