最小二乘支持向量机在模拟电路故障诊断中的应用

In this paper, .The principle component analysis (PCA) and least squares support vector machine (LS-SVM) are combined effectively in order to solve the problem of requiring many fault samples in fault diagnosis in analog circuit. In the process of fault diagnosis in analog circuit, first of all, calculate the voltage of some points on the circuit and collect original fault set, which is preprocessed by PCA to extract main components of the fault features in order to get the samples for training and testing. In the end, fault patterns under various states are classified using multi-class LS-SVM by the way of one against more. CV is presented to choose parameters of kernel function, which enhances preciseness rate of diagnosis .The simulation results show that the proposed method can detect and locate faults quickly and exactly.