Scholarlink

<p>There are thousands of high-speed trains traverse China’s railway network, generating vibrations in the railway roadbed and triggering seismic waves. The signals received by the Seismic sensor contain both seismic signals with narrow-band discrete spectral characteristics of the HSR source and noise signals with broad-band characteristics, resulting in more complex signals. In this paper, we propose a denoising method for seismic signals by combining the variational mode decomposition (VMD) with the K singular value decomposition (KSVD) learning dictionary algorithm. Firstly, the VMD is performed on the noisy HSR seismic signal to obtain a series of IMF components at different scales, and the noise-dominated IMF components are identified and removed by autocorrelation analysis. Then, the transitional IMF components are cumulatively reconstructed and subjected to secondary VMD, and the noise-dominated IMF components are again removed by autocorrelation analysis. Then, the remaining IMF components of the secondary VMD decomposition are superimposed with the remaining IMF components of the primary VMD to reconstruct a new signal, which is used as a training sample together with the original noise-containing signal, and the KSVD algorithm is applied to learn the sparse dictionary. Finally, the two noise-containing signals in the training samples are sparsely represented and the denoised seismic signals are reconstructed using the sparse coefficients to obtain the final denoised results. This method can provide technical support for the acquisition, processing and interpretation of seismic exploration data in complex surface areas, and help to improve the quality of seismic exploration data.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

High-speed Railway Signal Processing Based on a Combination of VMD and KSVD

  • 期刊论文

  • 发表日期：

    2025年7月26日

查看全文
  • Zi-ming Guo,
  • Zhi-yang Wang,
  • Yang Lei,
  • You-ming Li

摘要

There are thousands of high-speed trains traverse China’s railway network, generating vibrations in the railway roadbed and triggering seismic waves. The signals received by the Seismic sensor contain both seismic signals with narrow-band discrete spectral characteristics of the HSR source and noise signals with broad-band characteristics, resulting in more complex signals. In this paper, we propose a denoising method for seismic signals by combining the variational mode decomposition (VMD) with the K singular value decomposition (KSVD) learning dictionary algorithm. Firstly, the VMD is performed on the noisy HSR seismic signal to obtain a series of IMF components at different scales, and the noise-dominated IMF components are identified and removed by autocorrelation analysis. Then, the transitional IMF components are cumulatively reconstructed and subjected to secondary VMD, and the noise-dominated IMF components are again removed by autocorrelation analysis. Then, the remaining IMF components of the secondary VMD decomposition are superimposed with the remaining IMF components of the primary VMD to reconstruct a new signal, which is used as a training sample together with the original noise-containing signal, and the KSVD algorithm is applied to learn the sparse dictionary. Finally, the two noise-containing signals in the training samples are sparsely represented and the denoised seismic signals are reconstructed using the sparse coefficients to obtain the final denoised results. This method can provide technical support for the acquisition, processing and interpretation of seismic exploration data in complex surface areas, and help to improve the quality of seismic exploration data.