Automated Bayesian modal identification for full-scale transmission towers by comprehensive field tests

The modal identification is significant for the condition assessment of transmission towers. However, the effect of excitation patterns on the modal identification has not been adequately investigated for full-scale transmission towers. In this study, comprehensive field tests are conducted on a transmission tower. An automated Bayesian approach is proposed to identify the most probable values (MPVs) of the modal parameters with the posterior coefficient of variations (COVs). It is found that the identified MPVs of modal parameters by different excitation patterns are consistent, whereas the COVs show significant differences. The uncertainty of modal parameters by forced vibration test, despite with the highest SNR, is larger than that by ambient excitation. Moreover, ambient excitation is conducive to the identification of higher order modes for transmission towers, though the SNRs of acceleration responses are smaller. Detailed suggestions on the selection of field test schemes for transmission towers are further presented.