A dynamic Bayesian networks model for structural time-dependent reliability with deterioration

Z. Y. Wu, H. B. Sun, S. K. Liu

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

A dynamic Bayesian network (DBN) model is proposed for time-dependent reliability of structure with resistance deterioration. Gamma process is selected to model the deterioration process and loads are considered as random variables.The DBN of the structural time-dependent reliability is established by combining three models:deterioration model, reliability model, observation model, and is simplified with node elimination algorithm and discretization to allow exact inference. The framwork proposed is demonstrated by a one-bay frame with resistance deterioration. Comparison of the results obtained by DBN with Monte Carlo simulation approach, in terms of reliability index for 30 years shows the validity and accuracy of the DBN approach. Also exact inference of DBN such as filtering, prediction and smoothing with 10 years' evidence is presented in this paper.

Original languageEnglish
Title of host publicationSafety, Reliability, Risk and Life-Cycle Performance of Structures and Infrastructures - Proceedings of the 11th International Conference on Structural Safety and Reliability, ICOSSAR 2013
Pages1191-1197
Number of pages7
StatePublished - 2013
Event11th International Conference on Structural Safety and Reliability, ICOSSAR 2013 - New York, NY, United States
Duration: Jun 16 2013Jun 20 2013

Publication series

NameSafety, Reliability, Risk and Life-Cycle Performance of Structures and Infrastructures - Proceedings of the 11th International Conference on Structural Safety and Reliability, ICOSSAR 2013

Conference

Conference11th International Conference on Structural Safety and Reliability, ICOSSAR 2013
Country/TerritoryUnited States
CityNew York, NY
Period06/16/1306/20/13

Fingerprint

Dive into the research topics of 'A dynamic Bayesian networks model for structural time-dependent reliability with deterioration'. Together they form a unique fingerprint.

Cite this