针对非线性电阻网络,建立了规范式分段线性化故障诊断模型,提出了基于神经网络的非线性电阻网络硬故障诊断方法,从而克服了用一般方法对非线性电阻网络故障诊断中的的漏诊现象.
The paper presents method of canonical piece-wise linearization which can give all the multi-solutions in the nonlinear resistive network, and uses the backward-propagation neural networks for hard fault diagnosis of the nonlinear resistive network. Compared with the general methods, the proposed one overcomes the problem of fault missing location in the nonlinear resistance network.