基于BP神经网络的容差电路故障诊断  

作　　者：柏淑红[1] 陈宇峰[2] 

机构地区：[1]成都电子机械高等专科学校电子电气工程系 [2]成都电子机械高等专科学校电子电气工程系实验室

出　　处：《成都电子机械高等专科学校学报》2006年第4期4-7,22,共4页Journal of Chengdu Electromechanical College

摘　　要：模拟电路的可靠性决定了电子设备系统的可靠性。本文主要研究容差模拟电路故障诊断的神经网络方法,主要目的是解决模拟故障诊断中的容差问题和减少诊断时间。提出了基于BP网络的模拟电路故障诊断实现方法,其仿真结果达到了预期目的。The analog fault diagnosis research is one of the forefront of the testing area, one of the reason is that the reliability of electrical systems depends on that of analog circuits and systems . This paper focuses on the fault diagnosis of analog circuits with tolerance using the artificial neural networks method based on BP algorithm, The primary aim of the paper is to provide a mechanism to deal with the problem of element tolerance and reduce the testing time. The simulation results show that the proposed method can perform correct diagnosis in the analog circuits with tolerance.

关 键 词：故障诊断 容差电路 BP网络 

分 类 号：TP389[自动化与计算机技术—计算机系统结构]