A Time Series Based Damage Identification Approach to the Experimental Benchmark Studies
A time series based damage identification approach is presented aiming at the problem of damage diagnosis in structural health monitoring. The vibration signals obtained from sensors were firstly modeled as auto-regressive moving-average (ARMA) time series models, while a principal-component matrix derived from the AR parameters of these models was utilized to establish the Mahalanobis distance criterion functions. Then, a new index of damage-sensitive feature DSF was proposed. It was observed that the mean values of DSF had significant changed after damage. Thus, a hypothesis test involving t-test was further utilized to obtain a damage decision. To test the efficacy of this damage identification methodology, the approach had been applied to the ambient vibration tests of the IASC-ASCE Benchmark structure. Result shows that, the time series based index DSF is sensitive to structural subtle damage, and the proposed approach is able to identify all the damage scenarios introduced in the Benchmark structure.
Structural health monitoring Damage identification IASC-ASCE Benchmark Time series Damage-sensitive feature
Yi Liu Aiqun Li
College of Civil Engineering, Southeast University, Nanjing, China
国际会议
南京
英文
797-801
2007-10-16（万方平台首次上网日期，不代表论文的发表时间）