Stationarity-Based Approach for Lag Length Selection in Cointegration Analysis of Lamb Wave Data
Abstract
It is well known that when cointegration is used for the analysis of data in structural health monitoring applications, the choice of lag length has strong influence on damage detection results. The paper demonstrates how this problem could be solved. The solution utilizes the inversely proportional relationship between damage severity and stationarity of cointegration residuals. The method is validated using Lamb wave data from a structure exposed to temperature variations. The experimental results show that the proper lag length selection is essential and this can be achieved with the appropriate statistical analysis.
Origin : Files produced by the author(s)
Loading...