Abstract: As a core technology in the field of intelligent connected vehicles, environmental perception technology faces challenges in curriculum teaching such as outdated content, high practical training costs, and significant safety risks. Traditional virtual simulation technology, due to its lack of real-time interactivity, struggles to meet the demands of in-depth teaching. To address these issues, this study takes the course Environmental Perception System Assembly and Testing as a carrier, constructs a four-layer architecture model for industry-education integration based on digital twin technology, restructures the teaching content system of "virtual foundational cognitionsemi-physical system debugging", designs a practical training process of "task-driven–virtual pre-simulation–physical verification–iterative optimization", and establishes a school-enterprise co-constructed resource repository along with a multi-dimensional evaluation mechanism. Practice shows that students' course scores improved by 11.8%, and their willingness to participate in practical training increased by 62.5%, providing a referable practical pathway for industry-education integration and curriculum reform.

Key words: environmental perception; digital twin; industry-education integration; practical training

摘要： 环境感知作为智能网联汽车领域的核心技术，其课程教学面临内容滞后、实训成本高、 风险大等挑战。传统虚拟仿真技术因缺乏实时交互而难以满足教学深度需求。为此，以环境 感知系统装调与测试课程为载体，基于数字孪生构建产教融合四层架构模型，重构“虚拟基 础认知-半实物系统调试”的教学内容体系，设计“任务驱动－虚拟预演－实体验证－迭代优 化”的实训流程，并建立校企共建资源库与多维评价机制。实践表明，学生的课程成绩提升 了 11.8%，实训意愿增强了 62.5%，为产教融合和课程改革提供可参考的实践方案。

关键词: 环境感知；数字孪生；产教融合；实训