Key points are not available for this paper at this time.
Despite the potential benefits, industrial implementation of guided wave monitoring systems remains limited today. One of the reason behind this is the lack of consolidated reliability assessment techniques. Recent works clearly illustrated that the classical statistical methodology adopted from nondestructive evaluation to estimate POD is not directly applicable to structural health monitoring (SHM): the hypothesis of independence does not hold for the data obtained from permanently attached sensors and growing defects. The experimental determination of POD for SHM systems requires extensive and costly studies, in addition to the numerous representative structures. Involving numerical studies can complement experimental trials, inform POD evaluation and reduce this cost drastically. We propose in this work a comparative study to demonstrate the application of Model Assisted Probability Of Detection approach (MAPOD) for elastic guided wave monitoring systems. We apply classical and recently developed statistical methods proposed in the literature, using simulated data obtained thanks to the CIVA SHM module. The MAPOD estimation is validated in comparison to experimental studies.
Yılmaz et al. (Sat,) studied this question.