基于改进模糊聚类和支持向量机的模拟电路故障诊断

Analog Circuit Fault Diagnosis Based on Improved Fuzzy Clustering and Support Vector Machine

为解决神经网络训练需要大量的样本,且容易陷入局部最优,收敛速度慢等缺点,采用改进模糊聚类(IFC)和支持向量机(SVM)相结合的模拟电路故障诊断方法。利用小波分解技术提取待诊断电路的测试信息作为故障特征,在模糊聚类算法中为消除孤立点和噪声的影响,对不同样本点引入权值以提高聚类效果,结合改进的模糊聚类算法进一步降低故障特征的维数,将其作为支持向量机的输入量,进行模型训练并预测模拟电路的故障。仿真结果表明,此方法应用于电路故障诊断有效削减计算复杂度并提高了诊断精度。

An approach of improved fuzzy clustering（IFC） and support vector machine（SVM） for analog circuit fault diagnosis is proposed to solve the problems of excessive samples,local optimum and slow convergence of neural network training.Firsdy,testing information obtained from the analog circuit is used as fault feature by wavelet decomposition technology.Then,dimensions of fault feature are reduced by adding weight to samples of fuzzy clustering in order to eliminate influence of noise and isolated point and to improve clustering effect.Finally,it is imported to SVM to train model and predict the fault diagnosis of the analog aircuit.The simulation results show that the proposed method can reduce the calculation complexity and improve the diagnosis accuracy.