This study presents the development of a core digital twin algorithm for structural response prediction of the Container Ship, along with validation based on numerical simulations and model test data. A high-fidelity numerical model accounting for wave-induced structural behavior was first established. Based on this model, a physics-based reduced-order model(ROM) approach was applied to construct an efficient system for hydrodynamic–structural coupled analysis. Since the dynamic responses of ships and offshore structures strongly depend on the wave frequency characteristics, hydrodynamic pressure loads were efficiently applied to the reduced-order system using a spectrum-based formulation combined with interpolation, enabling real-time digital twin simulations. The proposed reduced-order system was designed to preserve the dominant dynamic characteristics of the full order model(FOM) while significantly reducing computational cost. In addition, a substructuring strategy was incorporated to allow model updating and prediction of localized structural responses associated with different degradation levels of ship components. The accuracy and reliability of the developed digital twin algorithm were validated through comparisons with FOM simulation results and backbone segmented model test data under both regular and irregular wave conditions. The results show that while the FOM requires an average computation time of approximately 496.33 seconds, the proposed reduced-order system predicts structural responses within about 1.3 seconds, achieving a computational efficiency improvement of approximately 383 times. With this substantial reduction in computational time, the reduced-order digital twin accurately reproduces the principal structural responses of the Container Ship, particularly the global bending moments, showing excellent agreement with model test measurements in both time and frequency domains. The proposed framework quantitatively confirms the effectiveness of physics-based reduced-order model with experimental data for digital twin applications. It provides a reliable foundation for real-time structural health monitoring and decision support during ship operation and can be extended to a wide range of marine and offshore structures.