부산대학교 도서관

Impact localization can provide accurate location information for rapid scanning and diagnosis of structural damage and sufficiently improve the efficiency of structural maintenance. The reference database method is commonly used to locate impact events, the key to which is how to match the incoming impact events with the reference impact events correctly by reliable features and valid metrics. However, for complex reinforced ribbed structures commonly used in aerospace, the stress wave generated by the impact is strongly coupled with the structure, which directly affects and interferes with the feature construction and similarity metric of the impact signal. In this article, the influence of complex structures on the accuracy of localization discrimination was overcome by using the Akaike information criterion (AIC) and wavelet packet transform (WPT). In the first step, the impact signals were first calibrated for time-of-arrival by the AIC function and then were reconstructed by WPT based on energy threshold to obtain reliable time–frequency characteristics. In the second step, the Pearson correlation coefficient was used to measure the correlation between two impact events, and then, the accurate impact coordinates were obtained by the centroid localization method based on the multiplexed sensing signals. The experimental validation results of low-velocity impacts on composite stiffened plate show that the method is able to locate the impact event with high accuracy under the working conditions of sparse sensing network and fewer reference samples, which meets the requirements of localization applications.