Wavelet entropy based damage identification using wireless smart sensors

Abstract

In this paper, a wavelet entropy based damage identification method is experimentally validated using wireless smart sensor units (Imote2) with TinyOS-based firmware. Recently, the wireless smart sensor network has drawn significant attention for applications in Structural Health Monitoring (SHM). Wavelet entropy is considered to be a damagesensitive signature that can be obtained both at different spatial locations and time stations to indicate changes in dynamic responses of structures. Compared to metrics based on the Fourier Transform, metrics based on wavelets require much simpler mathematics, with no complex numbers. Thus wavelet-based SHM methods would be easier to embed on motes. Wavelets can have other (mathematical) advantages when the structures are complex and the dynamic signals are non-stationary. Particularly, use of the relative wavelet entropy (RWE) has been extensively explored for use in damage detection using wireless smart sensors. First, sensor validation tests have been conducted using wireless and wired sensors. To verify an off-line time synchronization technique and the feasibility of using acceleration data from wireless sensors, modal identifications have been conducted using the ERA technique. Finally, the wavelet entropy based damage detection method has been demonstrated using Imote2 wireless smart sensors.