STRUCTURAL HEALTH MONITORING OF FASTENERS OF STRUCTURES AND WIND TURBINES FASTENERS USING MACHINE LEA
Structural Health Monitoring; Pattern Recognition; Vibrations; Machine Learning.
Mechanical and structural systems, such as rotating machines, airplanes, bridges, among others, are exposed to various types of efforts that, over time, induce wear and damage that affect the performance of the system or jeopardize its integrity. The present work aims to apply the method of Structural Health Monitoring (SHM) using Pattern Recognition and the vibration response of the structure to identify and monitor damage to fasteners (type of rivets, riveted joints, bolted, etc.). The work has a strong numerical-experimental character. From the measurements and data collection, the “vibration signature” of the system will be extracted with and without damage, which will be used together with the pattern recognition technique and Machine Learning (ML). For that, a classifier based on ML using frequency response function will be used in the work where it will be trained to process the extracted data and evaluate the structural integrity. Considering the scarcity of investigations aimed at the study of SHM using ML and pattern recognition for monitoring damage to fasteners, it is concluded that the current research project offers important contributions to the improvement of the proposed methodology, making it favorable in applications of wind turbines mechanical structures.