Real time model updating of stochastic nonlinear dynamic system using vibration data

• Main Author: Buntara, Fukent
• Other Authors: Cheung Sai Hung Joseph
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering::Civil engineering::Structures and design
• Online Access: http://hdl.handle.net/10356/67272
• Institution: Nanyang Technological University

Model updating is a branch of building health monitoring system in which building parameters such as its stiffness and damping components are updated after it sustained an external load or excitation. In this paper, the updating process focuses on three specific scenarios, which are dynamic excitation, non-linear behaviour of material, and real-time updating. The objective of this report is to find a method such that it can model a structure that under a given ground acceleration and the building acceleration data, the building parameters can be tracked and updated closely with the real structure behaviour. Newmark’s method is used to calculate the displacement, velocity and building acceleration given the prior parameters. Subsequently, the observed acceleration data is used to update the parameters, which will be used again for the calculation of the next time step and so on. The updating process is done every time step; hence it is called a real-time updating problem. Parameters updating process is done using stochastic approach, specifically the Bayesian state estimation technique. One of its method is called Particle Filter (PF) and this method is particularly good in dealing with non-linear system. The effectiveness of the numerical method and the updating process technique will be evaluated using two sample structures: a 5-DOF structure under sine wave excitation and a 7-DOF structure under El-Centro earthquake excitation.