Prediction error method-based second-order structural identification algorithm in stochastic state space formulation

Abstract

This paper presents a time-domain prediction error method (PEM)-based system identification method for the identification of second-order structural parameters of linear structural systems. The state space model of a linear structure is first expanded into linear AutoRegressive Moving Average with eXogenous excitation (ARMAX) or AutoRegressive Moving Average (ARMA) models by matrix fraction description. Second-order structural parameters are then estimated directly from measured time series data by applying the PEM to ARMAX and ARMA models. The proposed structural identification method can incorporate as much as possible structural information known a priori into the structural identification process to improve the identification accuracy. To avoid the practical difficulty often associated with input measurement, a two-stage structural health monitoring procedure is proposed. Once the mass matrix is identified from the first stage, data collected from ambient vibration survey can be used to identify other structural parameters in the second stage. The effectiveness of the proposed structural identification method as well as the corresponding structural health monitoring method is evaluated through numerical simulation in the context of the ASCE benchmark problem on structural health monitoring. The numerical results show that the proposed method is capable of locating and quantifying damages reasonably well in the presence of 10% measurement noise and limited sensor information. Copyright © 2006 John Wiley & Sons, Ltd.