Scholarlink

This study presents an approach for monitoring the transient variations in the multivariate coefficient of variation (MCV) using one-sided downward and upward control charts. Zero-crossing segmentation was employed to form subgroups, enabling efficient detection of process shifts. The methodology was applied to track geometry data collected from multiple railway sections across Hungary. Control charts were developed using both Maximum Likelihood Estimation (MLE) and Graphical Lasso (GLasso) techniques for estimating the precision matrix. Comparative analysis revealed that GLasso-based MCV charts with zero-crossing segmentation outperformed their MLE-based counterparts in detecting process changes.

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

Railway Track Quality Inspection via GLasso with Zero-Crossing Segmentation

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Abdaljbbar B. A. Dawod,
  • György Terdik

摘要

This study presents an approach for monitoring the transient variations in the multivariate coefficient of variation (MCV) using one-sided downward and upward control charts. Zero-crossing segmentation was employed to form subgroups, enabling efficient detection of process shifts. The methodology was applied to track geometry data collected from multiple railway sections across Hungary. Control charts were developed using both Maximum Likelihood Estimation (MLE) and Graphical Lasso (GLasso) techniques for estimating the precision matrix. Comparative analysis revealed that GLasso-based MCV charts with zero-crossing segmentation outperformed their MLE-based counterparts in detecting process changes.