Direct State-Space Models for Time-Varying Sensor Networks

Exploring the Efficiency of BIGDATA Analyses in SHM

“…The scalability and associated computational costs of an SHM procedure define the suitability for processing a very large dataset [11]. Many output-only SID algorithms [12][13][14][15][16][17][18] are not scalable procedures; their computational requirements typically increase cubically as with sensor channels and linearly with samples. Similar trends were observed for damage detection techniques using AR, SVR, and ARX models [11].…”