The optimization of steel structures encompasses a range of design tasks, including section and topology optimization, as well as the optimization of detailing and device parameters. For steel structures, section optimization is a representative task in which component sections are selected, adjusted, and verified through structural analysis and code-compliance checks. Traditional optimization algorithms have been widely used to identify feasible and economical designs, while surrogate-assisted methods have also been implemented to reduce the cost of repeated structural evaluations. However, these approaches are generally organized around case-specific searches rather than around the learning of reusable decision policies across related structural instances. Reinforcement learning (RL) provides acomplementary perspective by learning optimization strategies through interaction with structural environments. This review examines the use of RL for the intelligent optimization of steel structures, focusing on the section optimization of steel frames. First, the scope of the review is defined by organizing the main optimization tasks and forms of design variables in steel structures. Traditional optimization and surrogate-assisted methods are then reviewed as methodological foundations for single-case searches, constraint handling, and accelerated evaluation. Building on these foundations, RL-based studies are synthesized from the perspectives of structural-family formulation, environment-based policy learning, and generalization evaluation. The key barriers to developing code-compliant and engineering-oriented RL methods for steel-structure optimization are also discussed. This review provides a structured synthesis for understanding the role of RL in structural optimization and for developing more generalizable, verifiable, and engineering-oriented intelligent design methods.

Floating offshore wind turbines face growing integrity-management challenges caused by coupled corrosion and fatigue in harsh marine environments. Existing digital-twin frameworks are not yet well suited to combine multi-phase degradation physics with dynamic uncertainty quantification for this problem. To address this gap, this paper proposes a probabilistic digital twin framework that integrates sensor data acquisition, multi-physics simulation, and Bayesian inference for corrosion-fatigue prognosis. A three-phase damage evolution model is formulated to represent the transition from corrosion pitting to short-crack growth and long-crack propagation. Operational observations are assimilated recursively to update fatigue parameters and remaining useful life estimates. The framework is demonstrated using the IEA 15 MW reference wind turbine. The updated model identifies the onset of accelerated crack propagation at year 20 and reduces the 95% remaining-useful-life confidence interval from 40.4 years to 3.4 years. A maintenance strategy based on the updated failure probability reduces operational downtime by 58% and lifecycle cost by approximately 64.9% compared with a fixed-interval strategy. The results indicate that probabilistic updating can support more transparent inspection and maintenance decisions for floating offshore wind turbine structures under corrosion-fatigue degradation.

As a core material in modern construction, the early-age properties of concrete have a decisive impact on the safety and durability of civil engineering structures. However, systematic research on the mechanical properties of early-age concrete remains limited, particularly regarding the combined influence of cohesive and frictional properties on the material’s macroscopic mechanical behavior, which has not been thoroughly explored. To address this gap, this paper employs a decoupling method for testing the cohesion-friction mechanical properties of concrete, as proposed in previous work. This method successfully separates the cohesive and frictional properties of early-age concrete, validating its applicability under early-age conditions and obtaining typical failure modes following material performance degradation. Furthermore, by analyzing the evolution patterns of cohesive and frictional properties during deformation and strength development, the synergistic mechanism of cohesion-friction mechanical properties in early-age concrete was revealed. The results indicate that the responses of cohesive and frictional properties to hydrostatic pressure in early-age concrete exhibit significant differences. The reduction in macroscopic shear strength and stiffness is fundamentally attributed to the irreversible dissipation of cohesive strength. Ultimately, the mechanical behavior of early-age concrete gradually approaches that of granular materials without cohesion.