| 于阳1 王飞1 梁继文2.基于产业链技术体系的科学基金项目热点识别——以美国智能网联汽车产业为例[J].中国科技资源导刊,2026,(2):40~52 |
| 基于产业链技术体系的科学基金项目热点识别——以美国智能网联汽车产业为例 |
| Hotspot Identification of Scientific Fund Projects Based on Industrial Chain Technology System: A Case Study of the U.S. Intelligent Connected Vehicle Industry |
| 投稿时间：2025-11-24 |
| DOI： |
| 中文关键词: 美国政府基金;技术热点识别;产业链技术体系;智能网联汽车产业;BERTopic;双维分析 |
| 英文关键词: U.S. government funds, technical hotspot identification, industrial chain technology system, intelligent connected vehicle industry, BERTopic, two-dimensional analysis |
| 基金项目: |
| 作者 | 单位 | | 于阳1 王飞1 梁继文2 | (1. 江苏省科学技术情报研究所，江苏南京 210042
2. 苏州大学社会学院，江苏苏州 215127) |
|
| 摘要点击次数: 45 |
| 全文下载次数: 35 |
| 中文摘要: |
| 为解决已有技术热点识别研究中存在的产业链视角缺失、政府战略数据整合不足及跨学科技术边界模糊问题，研究构建基于三级产业链编码体系的技术热点解析框架，并基于美国智能网联汽车产业2014—2023年NSF(120672条)、SBA(59514条)资助项目，发现技术布局规律，为我国产业突破提供决策参考。在德尔菲法对产业链节点非独立性和协同性进行验证的基础上，融合GPT-4/Claude-3大模型及IEEE文献高频词构建领域自适应主题词库，运用BERTopic模型提取项目主题并以余弦相似度(阈值为0.7)实现节点匹配。基于产业链纵向、横向空间分布的双维分析，结合资助的强度，识别出3个技术热点:自动驾驶高精度定位算法、车联网通信协议优化技术及高可靠性传感器产业化技术。研究结论表明，我国需聚焦算法领域技术短板，强化传感器、算法与通信跨链协同，并通过动态技术清单机制规避车载娱乐系统等领域的资源错配。研究结果为攻克关键核心技术、优化产业创新政策提供了数据驱动的实证路径。 |
| 英文摘要: |
| To address the issues of missing industrial chain perspective, insufficient integration of government strategic data, and blurred interdisciplinary technological boundaries in existing technical hotspot identification studies, this study constructs a technical hotspot analysis framework based on a three-level industrial chain coding system. Using NSF (120 672 entries) and SBA (59 514 entries) funded projects in the U.S. intelligent connected vehicle industry from 2014 to 2023, it identifies technological layout patterns to provide decision- making references for China’s industrial breakthroughs. The study validates the independence and synergy of industrial chain nodes via the Delphi method, integrates GPT-4/Claude-3 large models with high-frequency terms from IEEE literature to build a domain-adaptive thesaurus, and applies the BERTopic model to extract project topics, achieving node matching with cosine similarity (threshold = 0.7) . Through two-dimensional analysis of vertical industrial chain and horizontal spatial distribution, combined with funding intensity, three technical hotspots are identified: high-precision positioning algorithms for autonomous driving, optimization technologies for vehicle networking communication protocols, and industrialization technologies for high- reliability sensors. The findings indicate that China should focus on addressing technical shortcomings in algorithms, strengthen cross-chain collaboration among sensors, algorithms, and communication, and avoid resource misallocation in areas such as in-vehicle entertainment systems through a dynamic technology inventory mechanism. This study provides a data-driven empirical pathway for breaking through “bottleneck” technologies and optimizing industrial innovation policies. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
