摘要

In this study an algorithm based on Shannon entropy, cross-time frequency spectrum (CTFS) and frequency varying velocities was proposed for structure health monitoring in two-layer plate. A linear array of two sensors is applied to capture the signals. By reducing the number of sensors we used a secondary pattern to get enough information for source locating. For this purpose a pattern for secondary points based on Shannon entropy and cost function was developed. Then to estimate the time delay of signals, cross-time frequency spectrum function was taken from captured signals. The time delay was calculated when CTFS function reached the maximum value. By taking short time Fourier transform of cross correlation function of captures signals and using dispersive curves, time delay and corresponding frequency dependent velocity are estimated. The experiments were carried out and the results showed high precision of presented algorithm.

  • 出版日期2017-9-1