Abstract
Ultrasonic Guided Waves are a highly effective tool for Structural Health Monitoring, offering both extensive coverage and versatility in detecting damage. However, due to external interference and information attack scenarios, the actual environment may be interspersed with strong noise, making accurate damage localization more difficult. This difficulty is further exacerbated if there is a lack of a priori knowledge about the noise distribution. To tackle this issue, this study proposes a novel framework that pioneers the application of Distributionally Robust (DR) optimization in SHM. The method does not require prior knowledge of noise distributions and effectively handles extreme noise conditions by leveraging perturbed sample spaces. Using a data-driven approach, ultrasonic guided wave signals are employed to enhance damage localization under complex and uncertain environments. Experiments were conducted on a 500mm × 500mm carbon fiber reinforced plastic plate with 20mm simulated damage. In valid area, the damage localization errors were 10.96mm and 18.26mm in two perturbed sample spaces at a signal-to-noise ratio of -1.78 dB and -2.04 dB.
Funding Information
  • National Key Research and Development Project (2023YFB3709605, 2022YFB4301300)
  • National Natural Science Foundation of China (62273206, 62073193, 62373219, 62373215)
  • Key Projects of the Ministry of Industry and Information Technology (TC220H057-2022)
  • Natural Science Foundation of Shandong Province (ZR2023QF023)
  • Key Technology Research and Development Program of Shandong Province (2021CXGC010204)

This publication has 46 references indexed in Scilit: