異常データ学習不要の統計的診断手法を用いた既存構造物ヘルスモニタリング(OS1-2 診断・解析,OS1 構造ヘルスモニタリング)

概要

論文の詳細を見る
This research is about a statistical diagnostic method for the structural health monitoring which is applicable to existing structures from the present moment. In this method, structural condition is statistically diagnosed by the change of the response surface. The response surface is a regression model of sensor's outputs. The change of the response surface is statistically tested with the F-test. In the F-test, a threshold of normal or fault condition is simply investigated with a theoretical F-probability distribution. Therefore, this diagnostic method only requires data of intact condition and does not require the complicated modeling and information of fault condition. To confirm the effectiveness of the presented method, the health monitoring of jet fans is conducted. In this study, load cells are mounted to a supporting system of the jet fan. The response surface is calculated from time-series load data measured by the load cells. As a result, the structural condition of the jet fan is successfully diagnosed.
2004-07-20