Health monitoring of mooring lines in floating structures using artificial neural networks

Health monitoring of mooring lines is essential to ensure the safe performance of floating structures during the service life. In the literature and offshore industries, damage diagnosis of mooring lines is based on fatigue analysis by considering rope behavior. Mostly, this type of diagnosis is accomplished by the results, obtained from the simulation model of mooring system. Further, one of the important factors in modeling is applying uncertainties in the simulation model. In this paper, due to the complex behavior of mooring lines, a new design of Radial Basis Function (RBF) neural network is proposed for damage diagnosis. Also, the modeling method is based on Rod theory and Finite Element Method (FEM). In the proposed modeling process, for improving the accuracy of the modeling, boundary conditions uncertainty are applied using Submatrix Solution Procedure (SSP). Additionally, round-off error is removed by SSP. Finally, the proposed modeling and diagnosis are investigated experimentally. The obtained results showed that proposed RBF has better performance compared with conventional one and other well-known methods in the literature.