复杂夹矸三维孪生煤层要实现对物理煤层的信息化描述，真实映射物理煤层的虚拟形态，构建模型是研究智能化采煤机高效截割的必要前提，同时也为实现“无人化”开采提供有效的数据保障。

以山东兖矿能源集团股份有限公司17层工作面采样数据为基础，研究了融合更多煤层地质结构条件信息的方法。采用不规则颗粒建模技术，孪生硫化铁结核；优化煤岩层颗粒填充技术；建立自定义接触模型模拟煤岩颗粒表面的粗糙度。以所构建的煤层三维孪生初始静态模型为基础，完成煤层应用插件的编译，根据地质特征为煤岩颗粒及结构体添加新属性实现孪生模型结构的修正；开展采煤机截割实验验证孪生模型的可行性。

结果显示：(1)实现了硫化铁结核、断层、顶底板构造的添加，完成煤层工作面三维孪生初始静态模型构建。(2)获得了可实现修正替换颗粒集的动态复杂夹矸煤层三维孪生模型。(3)采煤机截割实验的开展验证了其物理性能与实际煤岩的接近程度，截割复杂夹矸煤层三维孪生模型过程滚筒受载与实际煤岩之间的误差更小。(4)该模型在煤岩截割双向耦合试验中投入应用后，其优化后的重构煤层工作面三维孪生模型真实地反映了工作面煤层的形态，具有修正替换颗粒集功能的动态复杂夹矸煤层三维孪生模型可更准确快速地表达真实工作面的各种地质结构信息，提升煤岩截割数据信息获取的有效性。

3D digital twin for complex gangue-bearing coal seams are designed to achieve an information-based description of physical coal seams and faithfully map their virtual forms. digital twin is identified as a prerequisite for exploring the efficient cutting of intelligent shearers while providing an effective data guarantee for achieving unmanned mining.

Using data on samples from the mining face of coal seam 17 in the Yankuang Energy Group Co., Ltd. Shandong Province, this study proposed a methodology that integrates more information about the geological structural conditions of coal seams. Specifically, digital twining of iron sulfide nodules was performed using the irregular particle modeling technology, the coal-rock particle filling technology for coals was optimized, and the surface roughness of coal-rock particles was simulated using a user-defined contact model. Based on the constructed initial static 3D digital twin for a coal seam, this study compiled the application plugin for coal seams and added new attributes to coal-rock particles and structures according to geological characteristics, thus correcting the digital twin structure. The feasibility of the digital twin was verified using the cutting experiment with a shearer.

The results indicate that the initial static 3D digital twin for the coal-seam mining face was constructed by adding iron sulfide nodules, faults, and coal seam roofs and floors. The dynamic 3D digital twin for a complex gangue-bearing coal seam was established, enabling the correction and replacement of the particle set. The cutting experiment with a shearer demonstrates the similarity of physical properties between the modeled coal seam and actual coals and rocks, yielding minor errors in drum load during coal-rock cutting between the 3D digital twin for a complex gangue-bearing coal seam and actual coals and rocks. The 3D digital twin was applied to the bidirectional coupling test of coal-rock cutting, demonstrating that the optimized, reconstructed 3D digital twin for coal-seam mining face faithfully reflected the coal seam morphology of the mining face. The dynamic 3D digital twin for a complex gangue-bearing coal seam, allowing for the correction and replacement of the particle set, can reflect the information about various geological structures in actual mining face more accurately and rapidly, enabling more effective acquisition of coal-rock cutting information.