J0405-2-4 SI-F法におけるセンサ間の相関強度を用いた損傷検出限界の評価(知的材料・構造システム(2) 計測・損傷検出)

概要

論文の詳細を見る
The present study is about an unsupervised damage diagnostic method for the structural health monitoring named as SI-F method. In the proposed method, structural condition is statistically evaluated by detecting the change of relation between measurement of several sensors. The relation between sensor measurements are identified by the response surface methodology. The change of the response surface is statistically investigated with the F-test. In this method, the correlation model and the noise of fluctuation of absolute value affect accuracy of diagnosis of structure. Present study conduct a evaluation of the effect of regression error and scale of noise for the diagnostic accuracy via the SI-F method by simulated data and performing fatigue crack monitoring test via the SI-F method. As the result, the minimum R^2_<adj> value, which required to evaluate the damage, was analytically indicated, and the effectiveness was clarified from application to an experimental result.
2009-09-12