Time-dependent reliability analysis model under fuzzy state and its safety lifetime model

Yingshi Hu, Zhenzhou Lu, Jingyu Lei

Research output: Contribution to journalArticlepeer-review

9 Scopus citations

Abstract

In view of the lack of time-dependent reliability analysis model (TDRA) under fuzzy state, which is ubiquitous in engineering, a TDRA model under fuzzy state is proposed in this paper, followed by the corresponding safety lifetime model. To establish the TDRA model under fuzzy state, this paper firstly transforms the TDRA model under binary state into a form expressed by the time-dependent failure domain indicator function, and then the TDRA model under fuzzy state is derived based on the basic principle of the time-independent reliability analysis (TIRA) model under fuzzy state. By introducing an auxiliary variable and establishing the time-dependent generalized performance function, the TDRA model under fuzzy state is transformed into a generalized one under binary state, where the failure domain and safety domain are clearly defined. Then, the single-loop Kriging (SLK) surrogate model approach is used to efficiently estimate the time-dependent failure probability (TDFP) in the special service time interval under fuzzy state. Based on the generalized TDRA model and its efficient estimation, a safety lifetime model constrained by the target TDFP under fuzzy state and a corresponding efficient solving method are presented. Finally, examples are used to verify the rationality of the TDRA model and the safety lifetime model under fuzzy state established in this paper, and the efficiency of the algorithm is also validated.

Original languageEnglish
Pages (from-to)2511-2529
Number of pages19
JournalStructural and Multidisciplinary Optimization
Volume60
Issue number6
DOIs
StatePublished - 1 Dec 2019

Keywords

  • Fuzzy state
  • Safety lifetime model
  • Time-dependent reliability analysis model

Fingerprint

Dive into the research topics of 'Time-dependent reliability analysis model under fuzzy state and its safety lifetime model'. Together they form a unique fingerprint.

Cite this