Application of Independent Component Analysis in Structural Condition Identification

Abstract:

One kind of structure condition identification method based on Independent Component Analysis (ICA) is proposed in this paper. The change of structure condition is identified by the feature index extracted from structure vibration signals. It is effectual to identify the structure condition by transforming the test signals into the statistic characteristic space. The correlation coefficient is introduced to measure the correlative degree between the statistical characteristic of two structure conditions and the damage characteristic extraction index based on ICA is constructed. Considering the structure actual situation, assume the undamaged condition or small damaged condition as the benchmark condition. The analysis of structure vibration experiment indicate that the ICA characteristic index can identify the difference between damaged and benchmark condition and this method of ICA is practicable.

• Periodical: Applied Mechanics and Materials (Volumes 644-650)
• Pages: 4061-4065
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.644-650.4061
• Online since: September 2014
• Authors: Yang Yan*, Sheng Li, Xiao Long Jin
• Keywords: Benchmark Condition, Characteristic Extraction, Condition Identification, Independent Component Analysis

References

[1] P.C. Pandey, W.V. Barai: Multiplayer perceptron in damage detection of bridge structures. Computers & Structures, 1995, 54(4), pp.597-608.

DOI: 10.1016/0045-7949(94)00377-f

[2] K.F. Wang and X.D. Zhang: Research on damage identification in civil engineering structure, Highway, 2008, Vol. 1, No. 1, p.154–159.

[3] Q.B. He: Research on Multivariate Statistical Analysis for Machine Condition Monitoring and Diagnosis. Doctor's degree paper. Hefei, University of Technology of Science and Technology of China, (2007).

[4] Q.W. Zhang: Damage feature extraction and novelty detection for bridge health monitoring. Journal of Tongji University, 2003, 31(03), pp.258-262.

[5] A.J. Han, J. Xie: State-of-arts of health monitoring techniques for long span bridges, Bridge Construction, 2002(06), pp.69-73.

[6] Hyvärinen A, Karhunen J, Oja E: Independent component analysis. Publishing House of Electronics Industry, Beijing, (2007).