Premium
Damage identification in aircraft structures with self‐powered sensing technology: A machine learning approach
Author(s) -
Salehi Hadi,
Das Saptarshi,
Chakrabartty Shantanu,
Biswas Subir,
Burgueño Rigoberto
Publication year - 2018
Publication title -
structural control and health monitoring
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.587
H-Index - 62
eISSN - 1545-2263
pISSN - 1545-2255
DOI - 10.1002/stc.2262
Subject(s) - structural health monitoring , context (archaeology) , computer science , wireless sensor network , noise (video) , sensor fusion , identification (biology) , engineering , real time computing , artificial intelligence , botany , biology , computer network , paleontology , structural engineering , image (mathematics)
Summary Progress in self‐powered wireless sensor networks for structural health monitoring (SHM) have motivated the development of power‐efficient data communication protocols. One such approach is the energy‐aware pulse switching architecture, which employs ultrasonic pulses to transmit binary data through the material substrate. However, this technology creates time delays on the generated data due to the power budgets demanded for sensing and communication. The nature of data collected from such protocol thus requires the development of new analysis and interpretation methods. This study presents a robust damage identification strategy for aircraft structures, within the context of data‐driven SHM, using discrete time‐delayed binary data. A novel machine learning framework integrating low‐rank matrix completion, pattern recognition, k ‐nearest neighbor, and a data fusion model was developed for damage identification. Performance and accuracy of the proposed data‐driven SHM strategy was investigated and tested for an aircraft horizontal stabilizer wing. Damage states were simulated on a finite element model by reducing stiffness in a region of the stabilizer's skin. The reliability of the proposed strategy with noise‐polluted data was also validated. Further, the effect of variations in harvested energy on the performance of the approach was investigated. Results demonstrate that the developed machine learning framework can effectively detect the presence and location of damage based on time‐delayed binary data from a self‐powered sensor network.