doi:10.3850/978-981-08-7619-7_P044

ABSTRACT

The electromechanical impedance (EMI)-based structural health monitoring (SHM) comes to the forefront in the SHM community due to its practical potential for real applications. In the EMI technique, the selection of frequency bands is closely related to the accuracy of the results in assessing damages. It is important to select a proper frequency range sensitive to the expected damage type. The sensitive frequency range may vary to the type of the damage. Improper selection of the frequency range may lead to erroneous structural diagnosis. Therefore, more reliable and systematic approach for frequency range selection needs to be established. Thus, an innovative technique is proposed to find damage-sensitive frequency ranges autonomously using the neural network (NN) technique. The performance of the proposed NN approach has been examined for damage detection of loose bolts and cracks on a lab-scale aluminum plate. Then, the NN is combined with wireless impedance sensor node to detect real damages in a full-scale bridge. In results, it has been found that the proposed approach with wireless impedance sensor node can be successfully utilized to evaluate the damage type and severity in various damage cases.

Keywords: Piezoelectric sensors, Electromechanical impedance, Structural health monitoring, Neural networks, Frequency range selection, Wireless impedance sensor nodes.

© 2011 RPS