Scholarlink

<p>The worldwide implementation of Automatic Train Protection (ATP) plays a central role in railway safety, preventing overspeeding and also acting as inhibitor and mitigator of Signals Passed At Danger (SPADs). This study focuses on the Portuguese ATP system, evaluating its influence on SPAD occurrences and overall reliability. After introducing ATP functionality and relevant studies, multiple scenarios of ATP actions in SPAD events are analyzed to illustrate system variations based on train location and error occurrences. Two types of ATP errors are investigated: (i) On-Board System Errors and (ii) Balise Errors. For each type, a dedicated database is developed, incorporating explanatory variables. The study explores three Generalized Linear Models (Poisson, Hurdle, and Zero-Inflated Models) to assess error occurrences, comparing model performance using the Akaike Information Criterion (AIC) and selecting the best fit. The results highlight key factors affecting ATP reliability, revealing specific temporal and spatial dependencies, with both errors correlated with railway traffic and signaling density, and Balise Errors being more frequent during summer. By enhancing the understanding of ATP error patterns, this research contributes to the development of targeted mitigation strategies, ultimately improving railway safety and reducing SPAD risks.</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

Assessment of the automatic train protection system and its association to signals passed at danger in railway safety management

  • 期刊论文

  • 发表日期：

    2025年10月16日

查看全文
  • Guilherme Ribeiro,
  • Gonçalo F. Neves,
  • Miguel I. Grilo,
  • Virgínia Infante,
  • António R. Andrade

摘要

The worldwide implementation of Automatic Train Protection (ATP) plays a central role in railway safety, preventing overspeeding and also acting as inhibitor and mitigator of Signals Passed At Danger (SPADs). This study focuses on the Portuguese ATP system, evaluating its influence on SPAD occurrences and overall reliability. After introducing ATP functionality and relevant studies, multiple scenarios of ATP actions in SPAD events are analyzed to illustrate system variations based on train location and error occurrences. Two types of ATP errors are investigated: (i) On-Board System Errors and (ii) Balise Errors. For each type, a dedicated database is developed, incorporating explanatory variables. The study explores three Generalized Linear Models (Poisson, Hurdle, and Zero-Inflated Models) to assess error occurrences, comparing model performance using the Akaike Information Criterion (AIC) and selecting the best fit. The results highlight key factors affecting ATP reliability, revealing specific temporal and spatial dependencies, with both errors correlated with railway traffic and signaling density, and Balise Errors being more frequent during summer. By enhancing the understanding of ATP error patterns, this research contributes to the development of targeted mitigation strategies, ultimately improving railway safety and reducing SPAD risks.