To enhance the real-time response coverage and accuracy of structural health monitoring (SHM) systems for large ore carriers, this study investigates a stress-reconstruction method for a 32.5K DWT ore carrier under limited measurement data. A multi-level reconstruction framework is developed for both global and local responses: at the global scale, a full-ship finite element model is established under representative wave-load conditions, and time-domain stresses are reconstructed via the modal superposition method to obtain boundary conditions for the target cargo-hold segment. For local responses concentrated in the cargohold region, we propose a multi-load response superposition model that incorporates global– local coupling and multi-source combined loading scenarios. By decomposing structural responses under single-load cases and solving for load contribution coefficients using sparse sensor data, the model efficiently reconstructs stress fields in key structural zones without geometric simplification.Results show that the proposed multi-level approach yields reliable stress reconstruction across regions and operating conditions. The study thus establishes a scalable and interpretable method suited to complex hull structures, providing the theoretical basis and technical support for the engineering deployment of SHM systems on large ore carriers.