Research on reconstruction of signals for carbon fiber composite materials structural health monitoring based on compressed sensing

Abstract

The structural health monitoring technology (SHM) can use Lamb waves to monitor the health of structural parts online and in real time, so as to carry out safety assessment and life prediction of structural parts. As an important structural component of aviation, transportation and other fields, carbon fiber composite materials are prone to damage such as delamination, cracks, and fiber breaks during service. Therefore, it is necessary to use piezoelectric sensor arrays to excite Lamb waves to monitor carbon fiber composite materials actively to ensure the full operation of these important structural parts. In the process of monitoring, a higher sampling rate is usually used for data collection, which leads to a decrease in the speed of data transmission, storage, and processing. Therefore, it is necessary to compress the original data to reduce the amount of collected data. In this study, based on the compressed sensing technology, Gaussian random matrix is used to project the damage signal of Lamb wave for carbon fiber composite material into low-dimensional space, so as to obtain linear measurement value of sparse sampling and achieve the compressed sampling of signals. Finally, the reconstruction algorithm is used to realize the reconstruction of signals. Experiments show that the method of compressed sensing can compress and reconstruct Lamb wave signals, and it has good noise resistance. The absolute error of reconstruction is within [−0.3V, 0.3V], compressive sensing not only saves data storage space and improves data transmission speed, but also guarantees the accuracy of the reconstructed signal.