Optimization of Signal Pre-Processing for the Integration of Cost-Effective Local Intelligence in Wireless Self-Powered Structural Health Monitoring

Abstract

Recent research in Structural Health Monitoring (SHM) showed the ability of guidedwave based sensors networks to detect, localize and classify damage in its early stage. But, most of them still require the wiring of numerous devices. To avoid this technical restraint, particularly in airborne structures, wireless SHM system offer mass and cost savings, but powering the devices remains heavy. In this paper, actuators and sensors are powered by piezoelectric microgenerators, which harvest energy from the environing mechanical stress. The efficiency of the extraction process is optimized by a non-linear processing of the piezovoltage named Synchronized Switch Harvesting. Previous work showed that such techniques provide a stand-alone power source, whose performances meet the requirements of Wireless Transmitters and Receivers. Indeed, each sensing node has to feature its own power source in order to acquire its logical autonomy and thus, provide decentralized intelligence to SHM network. Although the diagnosis will be centralized, the amount of data passed to the central core of the network should be reduced to preserve a positive energy balance of the node. Various algorithms are compared in terms of sensitivity and computational cost, the latter directly impacting the consumption.