用支路撕裂法进行大规模模拟电路故障诊断的新算法

A New Approach for Fault Diagnosis of Large-Scale Analog Circuit by-Branch Tearing

本文首先介绍最近出现的一种用支路撕裂法进行的大规模模拟电路的故障诊断算法,讨论了该算法的容差问题。然后将故障预测技术、替代理论和迭加定理运用到支路撕裂法中,得到一个新的故障诊断方程,并给出了一种新的算法。还讨论了算法的并行处理问题。研究结果证明,当连续监测网络时,新算法的在线计算时间和容差问题均较原算法有很大的改进。

A approach for fault diagnosis of large analog circuit presented recently--the branch tearing is described firstly.The robustness problem of the approach is discussed. Then, a new approach for fault diagnosis using fault-prediction technique, substitution theorem, superposition theorem and branch tearing method is suggested and the parallel processing technique is discussed. The result shows that when the network is continuously monitored, the new approach results in a significant improvement in both computation and robustness over the original approach.