Online dual-rate decentralized structural identification for wireless sensor networks

In this paper, we propose a novel online dual-rate decentralized structural identification algorithm for wireless sensor networks. In this method, at the sensor nodes, the raw response measurements collected at each sensor are processed separately to obtain the preliminary local structural identification results. These results will then be compressed before transmitting with a much lower rate than the sampling rate to the central station for data fusion. At the central station, Bayesian fusion is utilized to integrate the compressed local identification results transmitted from the sensor nodes in order to obtain reliable global estimation. As a result, the large identification uncertainty in the local identification results can be substantially reduced. In addition to data compression, two different rates for sampling and transmission are used to alleviate the data transmission burden so that online model updating can be realized for wireless sensor networks. Examples using a 50-story shear-frame structure, a bridge, and a space truss are presented to demonstrate the performance of the proposed method.