摘要
针对线索表法在分析潜通路问题过程中约束条件过多的问题,在人工神经网络分析法的基础上应用网络流仿真进行潜通路分析.根据电路元件的电气特性以及人工神经网络的特点,建立了元件的定性仿真模型,确定电路网络的构成方式.通过网络流仿真模拟电流在电路系统中的扩散过程,预测电路中负载的响应.通过对比电路网络中负载的设计响应以及通过分析预测得到的负载响应,就可以判断出电路网络是否存在潜通路问题,并且找到发生潜通路问题的原因.该方法可以正确预测电路网络中的负载响应,克服了线索表法的缺陷,减少对分析已知条件的要求以及人为因素对分析结果的影响.
To solve the too many constraints problem in sneak circuit analysis of clue table method,network flow simulation was applied in sneak circuit analysis based on the research of artificial neural network(ANN) analysis method.Electrical elements' qualitative simulation models were established and circuit network's composition method was determined,based on electrical and ANN characteristics.The network flow simulation method simulated the current diffusion process and predicted the loads' responses in circuit network.The analysis results showed whether there was sneak circuit problem.Compared the loads' responses between the circuit network design and simulation analysis forecast,the difference was the sneak circuit problem.By using this method,the loads' responses in circuit network can be predicted correctly,the defects of the clue table method can be overcome,the analysis requirements of known conditions and the effects of human factors on the analytical results can be reduced.The sneak circuit problems in real circuit network can be solved effectively by this method.
出处
《北京航空航天大学学报》
EI
CAS
CSCD
北大核心
2012年第4期546-550,共5页
Journal of Beijing University of Aeronautics and Astronautics
关键词
潜通路分析
人工神经网络
网络流仿真
定性仿真
sneak circuit analysis
artificial neural network(ANN)
network flow simulation
qualitative simulation