An efficient approach for dynamic global sensitivity analysis of stochastic train-track-bridge system

Hua Ping Wan, Yi Qing Ni

Research output: Journal article publicationJournal articleAcademic researchpeer-review

17 Citations (Scopus)

Abstract

The parameters in time-varying train-track-bridge system (TTBS) are inevitably subjected to uncertainty, leading to variability in its dynamic responses. This work provides an investigation as to how uncertainty in the parameters influences the dynamic responses of time-varying TTBS, which refers to dynamic sensitivity analysis in the context of stochastic dynamic system. A powerful global sensitivity analysis (GSA), which maintains numerous attractive advantages over local sensitivity analysis (LSA), is proposed to measure the impact of uncertain system parameters on dynamic responses in a quantitative manner, thereby ranking them in order of importance. Furthermore, an analytical Gaussian process model (GPM)-based approach is developed to greatly alleviate the computational cost involved in dynamic GSA of time-varying TTBS. The proposed analytical GPM-based approach possesses the capacity to calculate the sensitivity indices of individual parameters as well as parameter clusters. The effectiveness of the proposed GPM-based dynamic GSA methodology is evaluated by comparison with the brute-force Monte Carlo simulation (MCS) on a mass-spring-damper system. The perfect agreement between the analytical GPM-obtained and MCS-derived time-varying sensitivity indices verifies the feasibility and reliability of the proposed methodology. An illustrative example is further provided to fully demonstrate the application of the developed GPM-based approach to dynamic GSA of time-varying TTBS. The dynamic GSA results enable us to gain insight into the temporal evolution of the global influences of individual parameters and groups of parameters on dynamic responses of TTBS.

Original languageEnglish
Pages (from-to)843-861
Number of pages19
JournalMechanical Systems and Signal Processing
Volume117
DOIs
Publication statusPublished - 15 Feb 2019

Keywords

  • Gaussian process model
  • Global sensitivity analysis
  • Parameter uncertainty
  • Probabilistic dynamic response
  • Train-track-bridge system (TTBS)

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Signal Processing
  • Civil and Structural Engineering
  • Aerospace Engineering
  • Mechanical Engineering
  • Computer Science Applications

Cite this