Smart monitoring of the health of marine structures based on artificial intelligence to identify scouring and sedimentation
Smart monitoring of the health of marine structures based on artificial intelligence to identify scouring and sedimentation
Payam PourSheikh Ali Asghari1 Alireza Mojtahedi2 Mohammed Ali Lotfollahi Yagin3
1) Misters
2) DOCTOR
3) DOCTOR
IOI Code
:
ABEB-ACAGAF
Publication :
4th International Conference on Recent Advances in Engineering, Innovation and Technology - Belgium(eitconf.com)
Abstract :
The phenomenon of scouring and sedimentation in the vicinity of foundations is one of the most critical safety challenges in the engineering of marine structures and bridges. This research, with a methodological and practical approach, deals with the development of an intelligent framework for structural health monitoring (SHM), which is based on the integration of digital twin technology and artificial intelligence. In this study, a multi-modal monitoring system is designed using distributed sensors such as optical fiber (DFOS) and piezoelectric systems to extract accurate data of substrate strain and changes. These data are used in the form of a digital twin model to continuously update the mechanical parameters and reduce the uncertainty of the finite element models. In the data analysis section, machine learning algorithms such as XGBoost have been used to predict the geometric changes of the bed, and the results show a very high accuracy of the model with an explanation coefficient exceeding 0.92 and a significant reduction in the RMSE error. Also, the effect of biologically inspired approaches in improving the sensitivity of sensors has been investigated. The findings of this research show that the transition from periodic inspections to intelligent and predictive monitoring not only greatly increases the accuracy of structural damage detection, but also ensures risk management and optimization of maintenance and repair costs in infrastructure structures by reducing the uncertainty of numerical models by 20 to 30 percent and life cycle costs by 30 to 45 percent.
Keywords : structure health monitoring marine structures machine learning scouring sedimentation