Correlated dual-layer degradation evaluation of ship hull anticorrosion coating systems using mode decomposition and Gaussian process regression

Ship hull anticorrosion coating systems, comprising surface coatings and metal substrates, are essential for protecting marine vessels against long-term corrosion in harsh ocean environments. However, existing degradation assessment models often treat coating and substrate deterioration separately, failing to capture their coupled degradation dynamics and limiting engineering applicability. To address this, we propose a correlated dual-layer degradation modelling framework tailored to ship hull coating systems. The framework integrates: (I) surface coating degradation modelling via Variational Mode Decomposition (VMD), the Mean Ratio (MR) method, and Gaussian Process Regression (GPR); (II) substrate corrosion estimation based on an equivalent circuit-driven delamination model and a nonlinear corrosion kinetics formulation; and (III) marine environmental sequence forecasting using a VMD-enhanced hybrid machine learning (VMD-HML) strategy. Validation under simulated marine conditions shows that the GPR model effectively captures coating degradation trends, while the proposed kinetics model accurately represents nonlinear substrate corrosion across various environments. The VMD-HML model outperforms conventional algorithms in environmental forecasting, enhancing long-term degradation prediction. This integrated framework provides a practical, scalable, and data-driven approach for evaluating the durability of ship hull anticorrosion coatings, offering valuable support for predictive maintenance, lifecycle management, and design optimisation of marine protective systems.