Scholarlink

<p>Reliability assessment of complex systems is challenging due to failure correlations among components and the difficulty of fusing reliability information across system levels. This paper proposes a copula-based multilayer Bayesian network method for reliability assessment. A nested copula model is used to capture dependencies and update marginal distributions layer by layer, enabling flexible modeling of correlated failures. The nested model is integrated with a multilayer Bayesian network to effectively combine system-level reliability information and enhance assessment accuracy. The method is validated on a binary-state hybrid system and a train traction drive system. Results show that the proposed CMBN approach overcomes limitations of traditional methods in high-dimensional dependence modeling and hierarchical reliability transfer, achieving higher accuracy and stronger engineering applicability. It provides an effective tool for the reliability assessment of complex systems such as rail transportation.</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

Reliability assessment method of complex systems with failure correlation and multi-level information fusion

  • 期刊论文

  • 发表日期：

    2026年5月11日

查看全文
  • Qing Xia,
  • Yonghua Li,
  • Zhenliang Fu,
  • Xiantao Zhang,
  • Shijia Liu

摘要

Reliability assessment of complex systems is challenging due to failure correlations among components and the difficulty of fusing reliability information across system levels. This paper proposes a copula-based multilayer Bayesian network method for reliability assessment. A nested copula model is used to capture dependencies and update marginal distributions layer by layer, enabling flexible modeling of correlated failures. The nested model is integrated with a multilayer Bayesian network to effectively combine system-level reliability information and enhance assessment accuracy. The method is validated on a binary-state hybrid system and a train traction drive system. Results show that the proposed CMBN approach overcomes limitations of traditional methods in high-dimensional dependence modeling and hierarchical reliability transfer, achieving higher accuracy and stronger engineering applicability. It provides an effective tool for the reliability assessment of complex systems such as rail transportation.