Wei Li

As a critical factor in ice load calculation for marine structures in cold regions, the growth mechanism and mechanical properties of rafted ice urgently require clarification. This study systematically investigated the growth patterns and flexural strength characteristics of rafted ice through laboratory-prepared specimens. Experimental results indicate that the thickness of rafted ice exhibits a negative correlation with both ambient temperature and initial ice thickness during growth. Due to the higher porosity of its frozen layer, the density of rafted ice decreases by approximately 8% on average compared to single-layer ice. Three-point bending tests demonstrate that, under the combined effect of high tensile strength in the lower ice layer and energy absorption by the porosity of the frozen layer, the flexural strength of rafted ice ranges from 1.12 to 1.34 times that of single-layer ice.

The randomness and complexity of ice loads present major challenges to the safety and stability of offshore platforms. Traditional methods for identifying ice loads often lack accuracy and adaptability under changing environmental conditions. This study proposes a novel inversion method based on Temporal Convolutional Networks (TCNs), integrating finite element simulation with deep learning to effectively identify random ice loads. A random ice load model is first developed, and its dynamic characteristics are validated through finite element analysis. The TCN model is then applied to capture the time-dependent features of ice loads. To improve the model’s generalization ability, its hyperparameters are optimized using particle swarm optimization (PSO). The results show that the TCN model achieves goodness-of-fit (R²) values of 0.821 and 0.808 on the training and test sets, respectively, indicating strong predictive performance. Under different ice thickness and velocity conditions, the model achieves R² values close to 0.99, demonstrating high robustness. This work represents the first application of TCN to ice load identification. By combining it with simulation data, we offer a high-precision, data-driven approach for dynamic load identification, enhancing the efficiency and reliability of safety assessments for conical offshore platforms.