Automatic operational modal analysis using statistical modelling of pole locations

Modal parameter extraction can be a time-consuming task. There are many applications where operator interaction is undesirable, for example for structural health monitoring. This paper presents an algorithm for automated operational modal analysis (AOMA) that utilizes one of the strong features of the stabilization diagram, namely the feature that physical modes tend to stabilize on the frequency axis for increasing model order. By employing statistical probability analysis and introducing a decision rule, that is based on the Modal Assurance Criteria (MAC), all modes in a user selected frequency range are automatically added to the output data-set. The algorithm has been tested on four different data-sets, a laboratory plexiglass plate, a suspension bridge, a medium-rise building and a ship. The results suggest that, the combined use of statistical probability analysis and a MAC based decision rule, successfully estimate modal parameters in an automated manner.