This study presents the development and validation of a complex-valued neural network model to predict wave conditions—such as wave direction, height, and period—surrounding floating structures. Accurate wave predictions are crucial for optimizing the design, operation, and maintenance of offshore platforms and floating offshore wind turbines, particularly in the context of digital twins. The proposed methodology leverages motion data obtained through numerical simulations of a floating structure to train the prediction model, enabling it to predict both the amplitude and phase information of the surrounding waves. The model successfully addresses the challenges of representing wave direction data in polar coordinates and capturing phase differences between motion components, which are difficult for traditional real-valued neural networks. The performance of the model was validated through various test cases, with the maximum prediction error found to be less than 10% and most predictions showing an error of less than 5%. Wave direction predictions demonstrated high accuracy, with errors consistently below 2%. While the model was trained using pseudo-measurement data, the results suggest that high-accuracy predictions can be achieved using real-world measurement data. This work contributes to enhancing wave prediction models for floating structures and is expected to improve the safety, performance, and long-term stability of such systems.

This study presents a hybrid machine learning framework combining an artificial neural network and a genetic algorithm to optimize chemical compositions of shielded metal arc weld metals for achieving targeted mechanical properties. First, a neural network model was trained using a large experimental database provided by Dr. Glyn M. Evans, which includes the chemical compositions and mechanical properties of over 950 shielded metal arc weld metals. The neural network model, optimized via Bayesian optimization, demonstrated high predictive accuracy for properties such as yield strength, ultimate tensile strength, and Charpy impact transition temperatures. To enable inverse design, a genetic algorithm-based optimization was applied to the trained neural network model, iteratively exploring the composition space to find optimal elemental combinations that match predefined mechanical property targets. The proposed hybrid approach successfully identified multiple feasible compositions that closely match the desired mechanical behavior, demonstrating the potential of neural network-assisted inverse design in welding alloy development.

Sloshing in liquid storage tanks is a critical phenomenon that affects the stability, performance, and safety of various engineering systems, including fuel tanks, offshore structures, and industrial storage units. The presence of internal structures, such as vertical baffles, significantly influences the natural sloshing frequencies and fluid motion. However, existing theoretical models often rely on simplified assumptions that restrict their capacity to capture the complexities of fluid–structure interactions in baffled tanks. This study integrates the coupled Eulerian–Lagrangian method with the impulse excitation technique to predict natural sloshing frequencies in a rectangular tank with vertical baffles. By analyzing the system’s response to an impulse excitation, we extracted the dominant sloshing frequencies while considering the impact of baffles on fluid dynamics. This computational approach provides a more realistic representation of sloshing phenomena and enables a parametric analysis of how various tank dimensions, fluid properties, and baffle configurations influence sloshing behavior. The findings of this study contribute to the improved design and optimization of liquid storage tanks, ensuring enhanced stability and performance in practical engineering applications. The integration of impulse excitation with the coupled Eulerian–Lagrangian method marks a significant advancement in sloshing analysis, offering a robust framework for understanding and mitigating the effects of sloshing in baffled tanks.

Sea level rise due to climate change poses an increasing threat to coastal ecosystems, infrastructure, and human settlements. However, accurately estimating sea level changes in regions without tide gauge observations remains a major challenge. While satellite altimetry provides wide spatial coverage, its accuracy diminishes near coastlines. In contrast, tide gauges offer high precision but are spatially limited. This study aims to develop an artificial neural network-based model for estimating relative sea level changes in coastal regions where tide gauge data are unavailable. Unlike conventional forecasting approaches focused on future time series prediction, the proposed model is designed to learn spatial distribution patterns and temporal rates of sea level change from a fusion of satellite altimetry and tide gauge data. A normalization scheme is applied to reduce inconsistencies in reference levels, and Bayesian optimization is employed to fine-tune hyperparameters. A case analysis is conducted in two coastal regions in South Korea, Busan and Ansan, using data from 2018 to 2023. The model demonstrates strong agreement with observed tide gauge records, particularly in estimating temporal trends of sea level rise. This approach effectively compensates for the limitations of satellite altimetry in coastal regions and fills critical observational gaps in ungauged areas. The proposed method holds substantial promise for coastal hazard mitigation, infrastructure planning, and climate adaptation strategies.