Statistical Moments of ARMA(n,m) Model Residuals for Damage Detection

摘要

In this study, a nonlinear auto-regressive moving average (ARMA) model residuals method is proposed for on-line structural damage detection. The residuals distribution of the ARMA(n,m) models are obtained for different m and n. An estimate of the ARMA model residual series standard deviation provides an accurate diagnosis of damage conditions. A repeatable threshold level that separates damaged from undamaged is detected. Additionally, statistical analysis applied to the datasets using the higher order moments that are more sensitive to disguised outliers. The results of one dynamic monitoring experiment are show that the developed method can detect damages with satisfactory precision.