通用大语言模型应用于理论力学辅助教学的探索与实践

EXPLORATION AND PRACTICE OF APPLYING GENERAL-PURPOSE LARGE LANGUAGE MODEL TO THE ASSISTED TEACHING OF THEORETICAL MECHANICS

  • 摘要: 人工智能技术为理论力学课程“难教、难学”的困境提供了新的解决途径。本文聚焦AI辅助教学的系统性实践,以DeepSeek通用大语言模型为工具,从解题辅助与动态可视化两个维度展开探索。在解题方面,利用AI分析典型力学问题的求解思路,并结合人工校正,显著提升了教师答疑效率。在可视化方面,借助AI生成基于HTML5的交互式动画,在一定程度上解决了传统动画资源固化、不可交互、制作成本高昂的问题。同时讨论了自然语言提示表述差异对AI辅助教学效果形成的现实约束。教学实践表明,AI辅助教学班在概念理解、期末成绩及课堂氛围方面均优于对照班级。AI技术不仅减轻了教师重复性劳动强度,也有助于降低优质交互资源的制作门槛,为缩小资源差距提供可能,并为理论力学课程的数字化转型提供了可推广的实践范式。

     

    Abstract: Artificial intelligence technology provides a new solution to the long-standing difficulties in teaching and learning the course of Theoretical Mechanics. Focusing on the systematic practice of AI-assisted teaching, this paper takes the general-purpose large language model DeepSeek as the research tool and explores two dimensions: problem-solving assistance and dynamic visualization. In terms of problem-solving, AI is employed to analyze the solution approaches to typical mechanics problems and, combined with manual correction, significantly improves the efficiency of teachers' responses to students' questions. In terms of visualization, AI is used to generate HTML5-based interactive animations, which to some extent addresses the problems of traditional animation resources, namely fixed content, lack of interactivity and high production costs. The practical constraints imposed on the effectiveness of AI-assisted teaching by differences in natural-language prompting are also discussed. Teaching practice shows that the AI-assisted class outperforms the control classes in conceptual understanding, final examination scores and classroom atmosphere. AI technology not only reduces teachers' repetitive workload but also helps lower the threshold for producing high-quality interactive resources, making it possible to narrow resource gaps and providing a replicable practical paradigm for the digital transformation of the Theoretical Mechanics course.

     

/

返回文章
返回