Bayesian operational modal analysis with asynchronous data, part I: Most probable value

- Simple model for random drift in asynchronous OMA data.
- Bayesian OMA formulation for asynchronous data.
- Fast algorithm for determining the most probable value of modal parameters using asynchronous OMA data.
- Verified with synthetic and laboratory data.

In vibration tests, multiple sensors are used to obtain detailed mode shape information about the tested structure. Time synchronisation among data channels is required in conventional modal identification approaches. Modal identification can be more flexibly conducted if this is not required. Motivated by the potential gain in feasibility and economy, this work proposes a Bayesian frequency domain method for modal identification using asynchronous 'output-only' ambient data, i.e. 'operational modal analysis'. It provides a rigorous means for identifying the global mode shape taking into account the quality of the measured data and their asynchronous nature. This paper (Part I) proposes an efficient algorithm for determining the most probable values of modal properties. The method is validated using synthetic and laboratory data. The companion paper (Part II) investigates identification uncertainty and challenges in applications to field vibration data.