Model updating is an important aspect of Structural Health Monitoring (SHM) because it ensures the accuracy and reliability of the structural models used to assess structural health. The need for model updating arises as a result of errors in the process of developing a theoretical model of a structure. Model updating allows models to be calibrated and validated using actual measured data, thereby improving their predictive capabilities. This paper implements a general Bayesian statistical framework for model updating. The Bayesian approach to model parameter updating entails solving a high-dimensional integral. It has been addressed by a numerical method known as the ‘Markov Chain Monte Carlo’ (MCMC) technique along with the ‘Metropolis–Hastings’ (MH) algorithm. To validate the framework, experimental data has been generated by applying dynamic base excitation to building prototypes using the shake table. The effectiveness of the proposed framework has been demonstrated by comparing theoretical response before and after model updating with the experimental response.

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

Vibration-Based Structural Health Monitoring Using Bayesian Model Updating

  • Nihal Khaimee,
  • Sonal Djanvijay

摘要

Model updating is an important aspect of Structural Health Monitoring (SHM) because it ensures the accuracy and reliability of the structural models used to assess structural health. The need for model updating arises as a result of errors in the process of developing a theoretical model of a structure. Model updating allows models to be calibrated and validated using actual measured data, thereby improving their predictive capabilities. This paper implements a general Bayesian statistical framework for model updating. The Bayesian approach to model parameter updating entails solving a high-dimensional integral. It has been addressed by a numerical method known as the ‘Markov Chain Monte Carlo’ (MCMC) technique along with the ‘Metropolis–Hastings’ (MH) algorithm. To validate the framework, experimental data has been generated by applying dynamic base excitation to building prototypes using the shake table. The effectiveness of the proposed framework has been demonstrated by comparing theoretical response before and after model updating with the experimental response.