题 目：时域可操作性模型分析：全贝叶斯方法

Time-domain operational modal analysis: a fully Bayesian approach

时 间：2017年12月25日上午9:00~10:00

地 点：7号楼2楼会议室

报告人：英国利物浦大学李宾宾博士

欢迎广大师生参加

                                  土木与交通学院

                                  2017年12月20日

报告人简介：

李宾宾博士2016年至今就职于英国利物浦大学风险与不确定性研究所，助理研究员职位，2016年在美国加州大学伯克利分校获得博士学位， 2012年和2009年在大连理工大学获硕士和学士学位。李宾宾博士研究兴趣包括：系统识别、贝叶斯统计、结构可靠性和蒙地卡罗取样。李博士现为美国土木工程学会（ASCE）、国际健康监测学会（ISHMII）和国际安全与可靠性学会（IASSAR）等学术组织会员，为Mechanical Systems and Signal Processing和 Structural Control and Health Monitoring等国际期刊审稿人。 

报告摘要：

     Operational modal analysis is the primary tool for modal parameter identification in civil engineering. Bayesian statistics offers an ideal framework for analyzing uncertainties associated with the identified modal parameters. However, the exact Bayesian formulation is usually intractable due to the high computational demand in obtaining the posterior distributions of modal parameters. In this talk, the variational Bayes method will be introduced to resolve the problem. Unlike the Laplace approximation and Monte Carlo sampling, the variational Bayes approach provides a gradient-free algorithm to analytically approximate the posterior distributions. The joint distribution of the state-transition and observation matrices as well as the joint distribution of the process noise and measurement error are calculated analytically using conjugate priors. The distribution of modal parameters is then extracted based on a first-order Taylor series expansion. A robust implementation of the method is discussed by using square-root filtering and Cholesky decomposition. The proposed approach is illustrated by its application to synthetic data, a lab model and the One Rincon Hill Tower in San Francisco.