Structural Response Evaluation of Krylov Subspace-Based Reduced-Order Model for Real-Time Structural Health Monitoring and Prediction of Container Ships

Recently, digital transformation has become crucial for the safe operation and extended lifespan of ships and offshore structures. Structural health management is gaining importance and driving interest in digital twin technology for monitoring structural integrity. Digital twins enable proactive maintenance through the real-time monitoring and prediction of structural responses. This study developed a reduced-order model (ROM) for a container ship and validated it against a full-scale model. To address irregular wave loading challenges, we applied a snapshot-based Krylov subspace method. Unlike conventional methods that use unit force vectors in axial directions, this approach considers time-varying pressure distributions from irregular waves. Numerical simulations showed that with 20 reduced orders, the structural response had a relative root mean squared error of 0.02% compared with the full-scale model, whereas computation time decreased by over 99%. The ROM maintained performance under varying heading angles and ship speeds, with 20–30 reduced orders balancing accuracy and efficiency. Thus, Krylov subspace-based model-order reduction is a valuable tool for predicting the structural responses of ships and offshore structures under real-time irregular waves. It is expected to be widely used in digital twins for structural health assessment, including damage detection and fatigue strength evaluation.