Data-driven condition monitoring of mooring systems: A demonstration with the M4 multi-float wave energy converter
Year: 2026
Type: article
Abstract: Monitoring the condition of mooring systems is essential for the safe operation and timely maintenance of floating platforms including wave energy converters (WECs). However, mooring dynamics are highly nonlinear due to nonlinearities in wave forcing, WEC dynamics, and floater-mooring coupling. These effects pose challenges in accurately identifying the condition of the mooring. In this study, we introduce a data-driven approach for condition monitoring that uses dynamic modeling to address these challenges. Linear and nonlinear models are compared to establish the dynamic relationship between motion and force for monitoring purposes, with the time-delay neural network (TDNN) demonstrating superior performance. For the TDNN, various training algorithms and activation functions are evaluated, with Bayesian regularization backpropagation and the hyperbolic tangent sigmoid (tansig) function producing the best results. Different motion inputs are considered, including surge (X) motion only, planar motion (XY), full spatial motion (XYZ), and six-degree-of-freedom (6-DOF) motion. The effectiveness of the proposed approach is demonstrated using two similar mooring system configurations with some distinct differences within a multi-float wave energy converter. The models are effectively trained on data from 22 wave conditions including highly directionally spread and crossing wave conditions, demonstrating generality of the model for complex ocean conditions. The model is subsequently tested against 16 wave conditions from an alternative configuration, where it effectively identifies a fault condition for all sea states, confirming the robustness and effectiveness of the method for condition monitoring applications.
Source: Ocean Engineering
Institution: University of Manchester
Language: English
Topic: Wave and Wind Energy Systems
Subfield: Ocean Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $4,150
OpenAlex ID: W7124988712