Identification of modal parameters of bridges considering temperature and wind effects

Li Shunlong,Li Hui,Ou Jinping,Li Hongwei
DOI: https://doi.org/10.3321/j.issn:1000-131X.2009.04.014
2009-01-01
Abstract:The dependence of natural frequencies, damping ratios on temperature and wind velocity are investigated. Because temperature and wind velocity measurement data may be closely correlated, the Nonlinear Principal Component Analysis (NLPCA) provides a powerful mathematical tool to transform the correlated variables (temperature and wind velocity) into a set of independent ones, the so called environmental principal components (PC), namely the temperature PC and the wind velocity PC. The investigation of temperature and wind velocity effects on modal parameters can then be transformed into the study of influences of environmental principal components. The NLPCA is also employed as a signal pre-processing tool to distinguish predominant environmental effects on structural modal parameters from other environmental factors. The pre-processed data by NLPCA includes implicitly the relationship between modal parameters and environmental PCs. The Artificial Neural Network (ANN) technique is employed to model the relationship between the pre-processed modal parameters and the environmental PCs. Numerical studies are carried out using the in-situ monitoring data of the natural frequencies, damping ratios, temperatures and wind velocity of a cable-stayed bridge. The results indicate that environmental factors dramatically impact the natural frequencies and the ANN technique is capable of modeling the relationship between the modal frequency and the environmental PCs. For some modes, the effects of environmental factors variations on modal damping ratios are observed, while the damping ratios are also affected by other factors besides temperature and wind velocity.
What problem does this paper attempt to address?