Abstract: Thermal-fluid-solid coupling refers to a complex multiphysics process involving the mutual interaction and dynamic feedback among the flow field, thermal field, and structural field during flight. It is of great significance to aerodynamic configuration design, thermal protection analysis, structural safety assessment, trajectory optimization, and flight control of hypersonic vehicles, aerospace planes, and reentry vehicles. Traditional approaches, including wind tunnel testing, theoretical analysis, and high-fidelity numerical simulation, can describe the underlying physical mechanisms with relatively high accuracy. However, under complex configurations, wide-range flight conditions, and strongly coupled multiphysics environments, these approaches still suffer from high experimental cost, long computational cycle, limited parameter coverage, and insufficient capability for rapid design iteration and online application. In recent years, data-driven methods have provided a new modeling paradigm for thermal-fluid-solid coupling by integrating wind tunnel data, numerical simulations, flight tests, and multi-source heterogeneous data. Existing studies have explored several key issues, including rapid prediction of complex aerodynamic characteristics, efficient reconstruction of local aerothermal loads, modeling of nonlinear structural responses, and reduced-order or surrogate modeling under multiphysics coupling conditions. These methods have shown great potential in improving computational efficiency, enhancing data utilization, and supporting rapid optimization, flight control, and virtual flight testing. Existing studies still mainly address single-field prediction and cross-field mapping, while unified models describing dynamic feedback among the flow, temperature, and structural fields remain limited. Based on this background, this review focuses on data-driven modeling of aerodynamic, aerothermal, and structural responses and their development toward thermal-fluid-solid coupling and systematically summarizes recent progress, typical methods, key challenges, and future trends from four aspects: intelligent aerodynamic modeling, intelligent aerothermal modeling, intelligent structural mechanics modeling, and data-driven applications of thermal-fluid-solid coupling. It distinguishes single-field surrogates, cross-field mappings, and feedback-coupled models, comparing their inputs and outputs, data requirements, use of physical information, and validation results. It also discusses multiple time scales, interface conservation, cross-field error propagation, long-term stability, and generalization across operating conditions. This review aims to provide a reference for multiphysics coupling modeling and intelligent analysis of complex flight vehicles.