Cantilever damage evaluation using impedance-loaded SAW sensor with continuous wavelet analysis and machine learning

Abstract

To monitor the health of large-scale structures, a wireless measurement system, such as a bridge, is required. One of the methods of monitoring the health of large-scale structures involves the application of an impedance-loaded wireless surface acoustic wave (SAW) sensor. Additionally, a pressure-sensor-loaded SAW sensor can detect the vibration of a cantilever. In this study, a continuous wavelet transform (CWT) is adopted to analyze the sensor responses. The CWT results obtained were classified into two categories based on the attenuation at each frequency, which include the exponential or linear type. Furthermore, machine learning was applied to evaluate cantilever damage. The results indicate that a high accuracy evaluation of damage is feasible with the proposed method.