基于Takagi-Sugeno型自适应模糊神经网络的模拟电路故障诊断

该文提出了一种基于Takagi-Sugeno型自适应模糊神经网络故障诊断方法。首先通过电路仿真获得故障样本,其次利用主成分分析对故障样本进行降维处理,减少自适应模糊神经网络的输入,降低训练时间,然后采用BP算法与最小二乘法相结合的混合学习算法训练自适应模糊神经网络的连接权值和隶属度函数。仿真结果表明,此方法能够快速有效地对模拟电路的故障进行诊断和定位,表现出了很好的应用潜力,在容差模拟电路故障诊断领域具有较好的应用前景。

Analog Circuit Fault Diagnosis Based on Takagi-Sugeno Type Adaptive Fuzzy Neural Network

This paper puts forward a fault diagnosis method based on Takagi-Sugeno type adaptive fuzzy neural network.First through the circuit simulation to obtain fault samples.Followed by the use of principal component analysis to reduce the dimension of fault samples,to reduce adaptive fuzzy neural network input,reduce training time.And hybrid learning algorithm composed of BP algorithm and least square method is used to adjust adaptive fuzzy neural network membership function parameter and connection weights.Simulation results show that this method is able to quickly and efficiently on analog circuit fault diagnosis and localization,showed good potential and has good application prospects in the field of analog circuit fault diagnosis.