Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4976029 | Journal of the Franklin Institute | 2007 | 11 Pages |
Abstract
A fault diagnosis method for analog circuit is proposed in this paper. An all-purpose amplifier is taken as the typical circuit under test (CUT). The chaotic signal, which is generated by an improved Chua's circuit, is employed as the excitation signal of CUT. The algorithm for phase-space reconstruction of chaotic time series is a combination algorithm of multiple autocorrelation and Î-test. The circuit state is estimated based on detecting the geometric change of Chua's attractor with a data-mining approach. For the purpose of information fusion, another eight features are extracted from the testing data to comprehensively determine the CUT states. A discussion is made for comparing the effectiveness of each feature according to the testing results.
Related Topics
Physical Sciences and Engineering
Computer Science
Signal Processing
Authors
Hong-Guang Ma, Xiao-Fei Zhu, Ming-Shun Ai, Ji-Dong Wang,