Chen Liang, Li Xing, Li Xun-bo, Huang Zuo-ying
School of Mechatronics Engineering, University of Electronic Science and Technology of China, Chengdu, 610054, People's Republic of China.
Rev Sci Instrum. 2009 Feb;80(2):025105. doi: 10.1063/1.3082021.
The commonly used and cost effective corrosion inspection tools for the evaluation of pipelines utilize the magnetic flux leakage (MFL) technique. The MFL signal is usually contaminated by various noise sources. In this paper, we propose that the pipeline flaw MFL signal is extracted using the ensemble empirical mode decomposition (EEMD) and the sparsity. At first, we introduce the EEMD method. The EEMD defines the true intrinsic mode function (IMF) components as the mean of an ensemble of trials, each consisting of the signal plus a white noise of finite amplitude. Moreover, sparsity selection restriction was defined. Then, The MFL signal is decomposed into several IMFs used for signal reconstruction. Some modes are selected to reconstruct a new signal considering their sparsity. Finally, the comparison is made with the empirical mode decomposition. At the same time, the comparison of the selection restriction between the sparsity and the energy is described. The results show that the EEMD and the sparsity is an efficient technology with the pipeline flaw extraction.
用于评估管道的常用且经济高效的腐蚀检测工具采用磁通量泄漏（MFL）技术。MFL信号通常会受到各种噪声源的污染。在本文中，我们提出使用总体经验模态分解（EEMD）和稀疏性来提取管道缺陷MFL信号。首先，我们介绍EEMD方法。EEMD将真正的固有模态函数（IMF）分量定义为一组试验的平均值，每个试验由信号加上有限幅度的白噪声组成。此外，定义了稀疏性选择约束。然后，将MFL信号分解为几个用于信号重构的IMF。考虑到它们的稀疏性，选择一些模态来重构一个新信号。最后，与经验模态分解进行比较。同时，描述了稀疏性和能量之间选择约束的比较。结果表明，EEMD和稀疏性是一种用于管道缺陷提取的有效技术。
