Efficient structural health monitoring with wireless sensor networks using a vibration-based frequency domain pattern matching technique

Abstract

Structural health monitoring (SHM) concerns the development of strategies to assess the condition of instrumented engineering structures. Most SHM techniques involve some form of physics-based modeling, which can be difficult and costly. We propose an efficient vibration-collecting method using wireless sensor networks (WSNs) that minimizes the need to perform modeling, making the SHM system easy to deploy on any structure with little to no modification. In our algorithm, we employ a two-layer stacking technique, similar to protocol stacking in networking. At the core of the algorithm is a frequency domain pattern matching technique applied to observed vibrations that is executed by each sensor node in the bottom layer of the algorithm. An energy management scheme for attaining the desired lifetime of the WSN is also proposed. The scheme is based upon a novel energy management circuit that decreases current for inactive sensor nodes, such to increase the lifetime of the nodes up to 90-fold. Upper bounds on system lifetime are determined experimentally using sensor nodes based upon the Zolertia Z1 mote. Finally, the pattern matching technique is validated on an experimental structure.