Experimental evaluation of uncertainty effect on real-time damage monitoring in prestressed concrete girders

Abstract

In this study, a new damage monitoring method using a set of parallel ANNs and acceleration signals is developed for alarming locations of damage in PSC girders. The problem addressed in this paper is defined as the stochastic process. In addition, a parallel ANN-algorithm using output-only acceleration responses is newly designed for damage detection in real time. The cross-covariance of acceleration-signals is selected as the feature representing the structural condition. Neural networks are trained for uncertain loading patterns and damage scenarios of the target structure for which its actual loadings are unknown. The uncertainty effect on real-time monitoring using the proposed method is evaluated from model tests on PSC beams for which accelerations were acquired before and after several damage cases.