Abstract: The precise analysis and efficient modeling of coal macromolecular structures are essential for understanding its macroscopic properties and microscopic reaction mechanisms, offering important theoretical insights for the clean and efficient use of coal. To comprehensively understand the current research status and development trends in coal macromolecular structure analysis and model construction, and to build molecular models that more closely resemble real coal systems, this study systematically discusses coal macromolecular structure from the perspectives of analytical methods, structural elucidation, and model construction. It discusses the limitations of current traditional construction methods, and looks forward to artificial intelligence （AI）-driven intelligent construction of coal macromolecular models. In the field of coal macromolecular structure research, analytical methods have evolved from early traditional methods relying on single spectroscopic approaches to modern integrated characterization systems that combine multi-scale and multi-technique strategies, enabling systematic revelation of the three-dimensional spatial structure and chemical bond distribution in coal. Based on this, studies have further clarified that coal macromolecules form a multi-scale complex system built on "basic structural units," consisting of regular and irregular components connected through covalent and non-covalent interactions, and have summarized the evolutionary patterns of coal structure with increasing coalification. In terms of model construction, the current commonly followed "experimental characterization - structural derivation - model construction" workflow faces bottlenecks such as poor cross-rank adaptability, low efficiency in large-scale model construction, and weak predictive capability for dynamic reactions, which hinder in-depth research on coal molecular structure and breakthroughs in clean and efficient coal utilization technologies. Accordingly, this paper systematically reviews the evolutionary trajectory of typical macromolecular models across different coal ranks to visually illustrate their development trends. Furthermore, it prospects the direction of AI-driven intelligent construction of coal molecular models, aiming to provide theoretical support for the intelligent and green transformation of the coal industry and to offer novel technical pathways for molecular-level precise regulation in the clean and efficient utilization of coal.