ABSTRACT
ABSTRACT This study aims to investigate a way to select sub-sensor groups for reduction of radio communications of whole system by confining radio communication within sensors inside sensor groups so as to save power of wireless sensor networks. Sparse-like system matrix is reconstructed for system identification utilizing the optimal sensor groups decided by correlations between sensors. The validity of the proposed approach is examined by utilizing the data both from simulations and field experiments on a steel girder bridge. Observations showed that system identification by reconstructing the sparse-like system matrix led to modal properties close to those from the original system matrix. Therefore, it encourages to utilize the sparse-like system matrix considering the optimal sub-sensor group for both system identification and damage detection.
