Data-fused method of fault diagnosis for analog circuits

Abstract

A data-fused fault diagnosis method based on wavelet packet decomposition of voltages of test nodes and current signals of the terminals stimulated is proposed in the paper. For the faults difficult to detect merely from voltages of test nodes, the current signals of the terminals which contain sufficient information with various faults are fused with the sampled node voltages of the circuit stimulated by the sources selected according to the principles proposed to make up for the insufficiency of the node voltages. This results in the maximization of feature vectors, more accurate classification of the faults and correct identification of the fuzzy sets of faults.