嵌入式数字电路瞬态故障自修复方法研究  

Research on Self-repairing Method of Transient Fault of Embedded Digital Circuit

作　　者：王银 Wang Yin(School of Mechanical and Material Engineering,Xi'an University of Arts and Science,Xi’an 710065,China)

机构地区：[1]西安文理学院机械与材料工程学院,西安710065

出　　处：《科技通报》2017年第9期154-158,共5页Bulletin of Science and Technology

摘　　要：为了以最快的速度恢复嵌入式数字电路的正常使用,需要对其瞬态故障自修复方法进行研究。提出一种基于小波包分解与神经网络相结合的嵌入式数字电路瞬态故障自修复方法,首先对采样的嵌入式数字电路电源的电流信号进行小波包分解和重构,提取嵌入式数字电路瞬态故障特征矢量;然后利用神经网络训练获得嵌入式数字电路瞬态故障诊断的输出结果;最后利用能力探测函数诊断嵌入式数字电路节点能量剩余情况,在整个嵌入式数字电路总体能量有限的情况下,适当地分配给嵌入式数字电路瞬态故障节点一部分能量,使其能够恢复正常工作,实现自修复。实验结果表明,本文方法能够减少自修复时间消耗。In order to use the fastest recovery of embedded digital circuit,the transient fault self repairmethod is researched.Propose a wavelet packet decomposition of embedded digital circuit transient faultcombined with neural network self repair method based on the first current signal sampling of embeddeddigital circuit power of wavelet packet decomposition fault feature extraction and reconstruction,embedded digital circuit transient output vector;then use neural network training for embedded digitalcircuit transient fault diagnosis;finally the capacity detection function for diagnosis of embedded digitalcircuit node residual energy,the total energy in the embedded digital circuit In limited circumstances,properly allocated to the embedded digital circuit transient fault node part of the energy,so that it canresume normal work,to achieve self repair.The experimental results show that this method can reducethe consumption of self repair time.

关 键 词：嵌入式数字电路 瞬态故障 自修复方法 小波包分解 神经网络训练 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]