Structural Health Monitoring Using Wireless Sensor Networks with Nonsimultaneous Sampling

Abstract

AbstractStructural health monitoring with wireless sensor networks (WSN) is an attractive alternative to traditional wired technology. The main challenges of WSNs are time synchronization, transmission of large amounts of data, and energy consumption. In this paper, autocovariance functions (ACFs) are estimated in all sensor nodes. Strict time synchronization is not necessary, because cross-correlations are not utilized. The measurement period must be long for a sufficient accuracy, but the number of samples in the transmitted ACFs is much smaller. The ACFs from all sensor nodes are transmitted to the base station for centralized data processing. Spatiotemporal correlation can be utilized, because for a stationary random process the ACFs have the same form as the free decay of the system. The covariance matrix is estimated using the training data from the undamaged structure under different environmental conditions. An extreme value statistics control chart is designed to detect damage. A numerical experiment was performed by simulating a bridge deck under stationary random excitation and variable environmental conditions. The excitation or environmental variables were not measured. Damage was a crack in a steel girder. Nonsimultaneous sampling of the WSN was simulated by selecting the starting time of the measurement randomly in each sensor node.KeywordsDamage detectionWireless sensor networkAutocovariance functionSpatiotemporal correlationTime synchronizationEnergy efficiencyEnvironmental effects