A novel algorithm for fault diagnosis of analog circuit with tolerances using improved binary-tree SVMs based on SOMNN clustering

In order to solving fault diagnosis of analog circuit with tolerances, noise, circuit nonlinearities and small sample sets, a novel multi-class classification algorithm which combined binary tree SVMs multi-classification based on self-organizing map nerve network (SOMNN) clustering roughly was proposed. The robustness characteristic of SOWN based on the separability between pattern classes and support vector machine (SVM) based on the theory of statistic learning for the small sample set were integrated in the algorithm. The SOWN was firstly applied to cluster layer by layer, by which structure of binary-tree SVMs multi-classifier for fault diagnosis was established, namely, the fault classes at each node of the tree were nailed down. Then according to the preprocess results of SOWN, SVM were utilized to segment each decision node accurately. The simulation results show us that compared with the several existent multi-class classification methods, the cut-rent algorithm has high accuracy and speed.