In this work, we investigate how to study the MHD activities in Tokamak plasma via the combination of singular value decomposition (SVD) and Hilbert–Huang transform (HHT) methods. We apply this approach to the Mirnov coil signal fluctuations analysis without any filtering technique. First, the principal axes (PAs) of a pick-up Mirnov signals are extracted by SVD analysis. Next, the harmonics of dominants PAs is obtained by empirical mode decomposition (EMD) analysis. Moreover, the time–frequency behavior of Mirnov signals are extracted by HHT. The proposed technique is employed to analyze Mirnov coils signals for mode type and frequency identification, especially in multimode MHD activities. We obtained Spatial–temporal structures of the Mirnov coils fluctuations in terms of correlation functions to better identification of mode number and frequencies of dominant MHD modes. We also present the results of this method applied to IR-T1 and Golem Tokamaks Mirnov coils signals. Consequently, satisfying results from SVD + HHT analysis method and spatial–temporal structures for IR-T1 and Golem Tokamaks Mirnov data observed.

中文翻译：

---

基于奇异值分解和希尔伯特-黄变换分析方法的托卡马克Mirnov线圈信号的MHD模式识别

在这项工作中，我们研究了如何通过奇异值分解 (SVD) 和希尔伯特-黄变换 (HHT) 方法的组合来研究托卡马克等离子体中的 MHD 活动。我们将此方法应用于 Mirnov 线圈信号波动分析，而无需任何滤波技术。首先，通过 SVD 分析提取拾取 Mirnov 信号的主轴 (PA)。接下来，通过经验模式分解 (EMD) 分析获得主导 PA 的谐波。此外，HHT 提取了 Mirnov 信号的时频特性。所提出的技术用于分析 Mirnov 线圈信号以进行模式类型和频率识别，尤其是在多模式 MHD 活动中。我们根据相关函数获得了 Mirnov 线圈波动的时空结构，以更好地识别主要 MHD 模式的模式数和频率。我们还展示了该方法应用于 IR-T1 和 Golem Tokamaks Mirnov 线圈信号的结果。因此，观察到 IR-T1 和 Golem Tokamaks Mirnov 数据的 SVD + HHT 分析方法和时空结构的令人满意的结果。