基于AI辅助的磁控溅射镀膜实验教学创新设计

Innovative Design of AI-Assisted Experimental Teaching for Magnetron Sputtering Coating

摘要: 磁控溅射技术已广泛应用于芯片制造、表面防护及高端模具涂层等高科技领域，开展磁控溅射镀膜实验教学可有效提高学生对PVD技术的了解。受限于设备数量、教师数量等教学资源限制，传统模式的教学下所取得的教学效果十分有限。本研究针对传统磁控溅射镀膜实验教学的不足，引入AI辅助教学手段，构建了“AI + 虚实结合课堂”的新型实验教学平台用于优化教学效果。该平台包含智能预习、动态优化及实时监控三大模块，通过虚拟仿真实验、智能参数优化及个性化学习反馈，有效解决了设备资源受限、试错成本高及教学监督不足等问题。研究结果表明，采用AI辅助教学的学生相比传统教学模式，在原理掌握、设备结构了解、前处理工序、镀膜过程性参数理解及课后思考等方面均有显著提升，总体教学效果提升13.5%。该创新设计不仅提升了学生的实验技能和创新能力，还为物理实验教学改革提供了新思路，展现了AI技术在工科教育中的广阔应用前景。

Abstract: Magnetron sputtering technology has been extensively utilized in high-tech domains, including chip fabrication, surface protection, and High-end mold coating. Implementing experimental instruction on magnetron sputtering coatings can significantly deepen students’ comprehension of Physical Vapor Deposition (PVD) technology. However, constrained by the limited availability of teaching resources, including equipment and qualified instructors, the teaching outcomes achieved under the traditional teaching model are significantly restricted. This study addresses the deficiencies in traditional magnetron sputtering coating experimental teaching by introducing AI-assisted teaching methods and constructing a novel experimental teaching platform, namely the “AI + virtual-real integrated classroom”, to enhance the teaching effectiveness. This platform is comprised of three key modules: intelligent previewing, dynamic optimization, and real-time monitoring. By leveraging virtual simulation experiments, intelligent parameter optimization, and personalized learning feedback mechanisms, it successfully addresses challenges such as limited equipment resources, high trial-and-error costs, and inadequate teaching supervision. The research findings indicate that students utilizing AI-assisted teaching demonstrate substantial improvements in various areas, including principle comprehension, equipment structure understanding, pretreatment procedures, coating process parameter interpretation, and post-class critical thinking, as compared to the traditional teaching approach. Additionally, the overall teaching effectiveness has been enhanced by 13.5%. This design enhances students’ experimental skills and innovation capabilities, offers insights into physics experiment teaching reform, and showcases the application potential of AI in engineering education.