ABSTRACT
This paper presents an output-only system identification method that is capable of processing new types of data such as BIGDATA, mobile sensor data, and others. Fixed sensor networks dominate those implemented for system identification (SID) and structural health monitoring (SHM). The rapid growth of smartphones has amplified the amount of digital data available and meanwhile, novel data collection techniques have been developed to provide more informative data. In the near future, new types of SHM data sets, not strictly those from a fixed sensor network, could be expected to become abundant in quantity and sufficient in quality, to the extent that they would provide a preferable alternative. Mobile sensor data and BIGDATA can be classified as dynamic sensor network (DSN) data and can be considered exactly with the truncated physical state-space model (TPM). This paper provides equations necessary to identify modal properties using data from this class. A compressed sensing SID strategy for BIGDATA is presented and applied to ambient accelerations recorded at the Golden Gate Bridge. The compressed data contain dense spatial information in a compact size and provide accurate modal estimates.
