中文 | English

J. Hefei Univ. Tech. (Nat. Sci.)

合肥工业大学学报(自然科学版)

JOURNAL OF HEFEI UNIVERSITY OF TECHNOLOGY (NATURAL SCIENCE)

Competent Authorities: Ministry of Education of the People's Republic of China

Sponsor by: Hefei University of Technology

ISSN 1000-1646, CN 34-1001/N

Single-channel blind source separation algorithm for eddy current nondestructive testing signal

Download Full Text

单通道涡流无损检测信号盲源分离算法

Journal Information

J. Hefei Univ. Tech. (Nat. Sci.), Feb.2023, Vol.46 No.2: 175-179, 267

DOI: 10.3969/j.issn.1003-5060.2023.02.006

Authors

YANG Zhiwei, NAN Xinyuan

(School of Electrical Engineering, Xinjiang University, Urumqi 830047, China)

Abstract and Keywords

Abstract: Aiming at the aliasing problem of single detection signal in pulsed eddy current testing (PECT) system, a single-channel blind source separation algorithm based on empirical mode decomposition and fast independent component analysis (EMD-FastICA) is proposed. The proposed algorithm separates the mixed signal by EMD, and then estimates the number of source signals by singular value decomposition (SVD). According to the estimated number, the detection signal and the intrinsic mode function (IMF) components are combined to get the new virtual signal. Finally, FastICA is utilized to get the estimate of source signals. According to finite element simulation, the algorithm can effectively separate single-channel mixed detection signal and its performance is better than that of the single-channel blind source separation algorithm based on wavelet decomposition.

Keywords: independent component analysis (ICA); empirical mode decomposition (EMD); pulsed eddy current testing (PECT); single-channel blind source separation; finite element simulation

作者信息

杨智伟，南新元

（新疆大学电气工程学院，新疆乌鲁木齐 830047）

摘要和关键词

摘要: 针对脉冲涡流无损检测(pulsed eddy current testing, PECT)系统中获取单一检测信号存在的混叠问题,文章提出一种基于经验模态分解(empirical mode decomposition, EMD)和快速独立分量分析(fast independent component analysis, FastICA)的单通道盲源信号分离算法。该算法首先通过EMD对混合观测信号分解,然后利用奇异值分解(singular value decomposition, SVD)估计源信号数目,根据估计得到的源信号数目将观测信号和对应模态分量构成新的虚拟信号,最后利用FastICA算法分离得到源信号的估计。有限元仿真实验表明该算法能有效分离单通道混合检测信号,并且优于小波分解的单通道盲源分离算法。

关键词: 独立分量分析(ICA); 经验模态分解(EMD); 脉冲涡流无损检测(PECT); 单通道盲源分离; 有限元仿真

Funding

国家自然科学基金资助项目（61463047）

Previous Next

Back to Contents