主办：陕西省汽车工程学会
ISSN 1671-7988  CN 61-1394/TH
创刊：1976年
汽车实用技术

汽车实用技术 ›› 2026, Vol. 51 ›› Issue (12): 114-120.DOI: 10.16638/j.cnki.1671-7988.2026.012.020

• 汽车教育 • 上一篇    

汽车系统动力学研究生课程知识图谱构建与 可视化交互系统研究

张智明，胡光旭，赵治国*，甄玉君，赵欣，符鹏鹏   

  1. 同济大学 汽车学院
  • 发布日期:2026-06-23
  • 通讯作者: 赵治国
  • 作者简介:张智明（1979－），男，博士，副教授，研究方向为车用燃料电池关键技术 通信作者：赵治国（1971－），男，博士，教授，研究方向为新能源汽车动力系统
  • 基金资助:
    同济大学研究生课程“知识+能力”图谱建设项目《车辆系统动力学》（2024KCTP23）；同济大学 2025 年研 究生知识微单元-智能电动车辆项目（1700106041）；2025 年同济大学研究生教育研究与改革重点项目 （2025ZD05）

Research on Knowledge Graph Construction and Visual Interactive System for the Graduate Course Vehicle System Dynamics

ZHANG Zhiming, HU Guangxu, ZHAO Zhiguo* , ZHEN Yujun, ZHAO Xin, FU Pengpeng   

  1. School of Automotive Studies, Tongji University
  • Published:2026-06-23
  • Contact: ZHAO Zhiguo

PDF

摘要： 针对车辆工程研究生核心课程汽车系统动力学存在的公式推导繁杂、知识点耦合度高、 逻辑关系隐蔽等教学痛点，文章提出全流程课程知识图谱构建与可视化交互系统开发方法。 该研究基于 Python 与 Neo4j 图数据库技术，构建包含 300 余个知识实体、900 余条语义关系的 细粒度课程知识库，重点阐述多源异构数据提取、多策略知识融合及前后端分离可视化交互的 关键技术实现，同时独立开发后台管理系统保障图谱动态更新。在同济大学 2024 级该课程研 究生教学中的应用验证显示，该系统有效提升两个班 80 名学生的知识体系构建效率，实验组 考核成绩较对照组提升 12.5%，课程满意度达 95%，为研究生智慧课程高质量建设提供了可 推广范式。

关键词: 汽车系统动力学；知识图谱；知识融合；可视化交互；智慧教学

Abstract: Aiming at the teaching difficulties of cumbersome formula derivation, high correlation of knowledge points and hidden logical relations existing in Automotive System Dynamics, a core course for postgraduates majoring in vehicle engineering, this paper proposes a full-process method for constructing curriculum knowledge graphs and developing visual interactive systems. Based on Python and Neo4j graph database technology, this study establishes a fine-grained curriculum knowledge base with more than 300 knowledge entities and over 900 semantic relationships. It mainly illustrates the key technical implementations of multi-source heterogeneous data extraction, multi-strategy knowledge fusion and front-end and back-end separated visual interaction, and independently develops a background management system to ensure dynamic updating of knowledge graphs. Application verification in the teaching of this course for 2024-grade postgraduates at Tongji University proves that the system effectively improves the efficiency of knowledge system construction among 80 students from two classes. The examination scores of the experimental group rise by 12.5% compared with the control group, and the course satisfaction rate reaches 95%. It provides a popularizable model for the high-quality construction of intelligent postgraduate courses.

Key words: Vehicle System Dynamics; knowledge graph; knowledge fusion; visual interaction; smart teaching