Distributed Decision-Making in Wireless Sensor Networks for Online Structural Health Monitoring

Qing Ling,Zhi Tian,Yue Li
DOI: https://doi.org/10.1109/jcn.2009.6391348
2009-01-01
Journal of Communications and Networks
Abstract:In a wireless sensor network (WSN) setting, this paper presents a distributed decision-making framework and illustrates its application in an online structural health monitoring (SHM) system. The objective is to recover a damage severity vector, which identifies, localizes, and quantifies damages in a structure, via distributive and collaborative decision-making among wireless sensors. Observing the fact that damages are generally scarce in a structure, this paper develops a nonlinear 0-norm minimization formulation to recover the sparse damage severity vector, then relaxes it to a linear and distributively tractable one. An optimal algorithm based on the alternating direction method of multipliers (ADMM) and a heuristic distributed linear programming (DLP) algorithm are proposed to estimate the damage severity vector distributively. By limiting sensors to exchange information among neighboring sensors, the distributed decision-making algorithms reduce communication costs, thus alleviate the channel interference and prolong the network lifetime. Simulation results in monitoring a steel frame structure prove the effectiveness of the proposed algorithms.
What problem does this paper attempt to address?