During the last decades, the Operational Modal Analysis (OMA) techniques for the dynamic identification have reached a wide application in the large civil structures. Then the focus has been moved on the introduction of the Automated OMA (AOMA) techniques as a component of a wider Structural Health Monitoring (SHM) framework. On the one hand, the automatization of the process is needed to extract the modal properties from huge database of measurements. On the other hand, the methods should be able to minimize the model bias, avoiding the spurious and mathematical modes. Moreover, all the fix thresholds defined a-priori should be carefully set to avoid some tracking errors. This paper presents a new AOMA for the extraction of the modal parameters from long-term monitoring data. The methodology herein discussed aims to minimize the errors that can arise during the modal identification. Moreover, the information obtained during an observation period have been used to define suitable adaptive thresholds. For illustrative purposes, the methodology has been applied to the data acquired from long-term monitoring systems installed on two masonry towers subjected to different operational conditions. The effectiveness of the proposed method has been checked on the extracted modal parameters
展开▼
