Condition Identification of Bolted Joints Based on Autoregressive Model

Autor: Zhang, Yan, Zeng, Guo Ying, Zhao, Deng Feng, Li, Ming Yan
Zdroj: Advanced Materials Research; January 2012, Vol. 433 Issue: 1 p617-621, 5p
Abstrakt: The difference signals, which between upper and lower shells of flange bolted joints structure, was applied to establish the autoregressive model for system condition identification. Firstly, an AR model is built by difference signals. The established AR model is used as a filter to process the difference signal in test state under the same condition and output residual series. Then the statistical parameters, such as Itakura distance, skewness, kurtosis and variance, are used to handle residual series. The results of experiment show that Itakura distance is a useful eigenvalue to identify the bolted joints condition.
Databáze: Supplemental Index