融合人工智能的《现代表面技术》中化学气相沉积虚拟仿真实验教学方案设计

Design of a Virtual Simulation Experimental Teaching Scheme for Chemical Vapor Deposition Integrating Artificial Intelligence in Modern Surface Technology

摘要: 人工智能技术的快速发展为工程类核心课程的教学方式提供了新的变革契机。《现代表面技术》课程内容涉及多种表面处理工艺与表征方法，传统以教师讲授与静态演示为主的教学模式难以充分支撑学生对复杂工艺机理的理解与综合应用能力的培养。为此，本文在梳理相关研究基础上，聚焦于课程中的化学气相沉积(Chemical Vapor Deposition, CVD)虚拟仿真实验模块，提出一套融合人工智能技术的教学方案设计思路。方案以基于知识图谱的个性化学习路径推荐和基于学习行为数据的学习分析为核心技术支撑，通过构建CVD工艺知识图谱、设计虚拟仿真实验场景以及规划“课前引导–在线实验–结果反思–工程情境拓展”的教学流程，为学生提供分层递进、因材施教的学习路径。同时，从教学评价与教育伦理视角出发，构想了融合过程性学习数据的多元评价框架，并探讨了在新型教学模式下教师从“知识传授者”向“学习设计者与数据解读者”的角色转变。本文旨在为《现代表面技术》课程的智能化升级提供一种可借鉴的方案设计与路径探索，为后续开展实证研究奠定方法基础。

Abstract: The rapid development of artificial intelligence (AI) has brought new opportunities for transforming the teaching of engineering core courses. Modern Surface Technology covers a wide range of surface treatment processes and characterization methods, and traditional lecture-based and demonstration-based teaching models are insufficient for supporting students in understanding complex process mechanisms and developing comprehensive application abilities. Based on a review of related studies, this paper focuses on the chemical vapor deposition (CVD) virtual experiment module in the Modern Surface Technology course, and proposes an AI-enhanced teaching scheme at the stage of conceptual design. The scheme is mainly supported by a knowledge-graph-based personalized learning path recommendation mechanism and learning analytics based on students’ behavioral data. By constructing a CVD process knowledge graph, designing immersive virtual experiment scenarios, and organizing a “pre-class guidance, online experiment, reflection on results, engineering context extension” teaching flow, the scheme aims to provide a layered and differentiated learning pathway for students. In addition, from the perspectives of assessment and educational ethics, a multi-dimensional evaluation framework integrating process data is outlined, and the evolving role of teachers as learning designers and data interpreters in AI-supported teaching is discussed. This work is positioned as a scheme design and pathway exploration, which is expected to provide methodological support for subsequent empirical studies.