Feature extraction of analogue circuit fault signals via cross‐wavelet transform and variational Bayesian matrix factorisation. Issue 2 (1st March 2019)

• Record Type: Journal Article
• Authors: He, Wei; He, Yigang; Li, Bing; Zhang, Chaolong
• Is Part Of: IET science, measurement & technology. Volume 13:Issue 2(2019)
• Page Start: 318
• Page End: 327
• Publication Date: 2019-03-01
• Subjects: optimisation -- fault diagnosis -- feature extraction -- time‐frequency analysis -- analogue circuits -- Bayes methods -- support vector machines -- wavelet transforms -- matrix decomposition -- electronic engineering computing -- signal representation; analogue circuit fault signals -- analogue circuits -- XWT -- VBMF -- joint time‐frequency representation -- time‐frequency information -- nine‐dimensional feature vector -- cross‐wavelet transform -- feature extraction -- variational Bayesian matrix factorisation -- flower pollination -- support vector machine; Measurement -- Periodicals; Electrical engineering -- Periodicals; Electronics -- Periodicals; Nanotechnology -- Periodicals; Electromagnetism -- Periodicals; Medical instruments and apparatus -- Periodicals; 621.3
• DOI: 10.1049/iet-smt.2018.5432
• Languages: English
• ISSNs: 1751-8822

Abstract

Abstract : Analogue circuits are one of the most commonly used components in industrial equipment, but circuit failure may lead to significant causalities and even enormous financial losses. To address this issue, in this work the authors propose a new feature extraction scheme based on cross‐wavelet transform (XWT) and variational Bayesian matrix factorisation (VBMF). Primarily, fault signals acquired from defect circuits are collected and processed by using XWT to obtain the joint time‐frequency representation. VBMF is utilised to fetch the time‐frequency information of the fault signal. A nine‐dimensional feature vector is then constructed. Finally, a support vector machine optimised by a flower pollination algorithm is introduced to locate faults. Results show that the proposed approach can effectively locate the different kinds of defection while achieving a higher accuracy.