现代纺织技术

• •    

"基于LDA算法的关键共性技术识别研究 ——以智能纺织领域为例"

  

  • 网络出版日期:2025-03-09

Research on the identification of key common technologies based on LDA algorithm: A case study of the intelligent textile industry

  • Online:2025-03-09

摘要: 关键共性技术研究是推进国家制造业创新发展的重要支撑,准确识别关键共性技术成为推动产业转型升级的一大助力。以专利数据为基础,运用LDA主题模型挖掘出隐藏的高强度技术主题,通过共现率指标评估技术主题的共现程度,以此归纳出共性技术。在此基础上,运用网络分析方法,结合度中心性、接近中心性以及结构洞3个拓扑指标将各技术主题的关键程度进行量化分析,进一步识别出关键共性技术。研究发现:智能纺织领域的关键共性技术主要包含在工艺材料、工艺流程和工艺设备中;石墨烯纤维、柔性纳米纤维、聚氨酯整理剂等是工艺材料中的关键共性技术;智能缝纫技术、超疏水涂层技术、水凝胶导电技术等是工艺流程中的关键共性技术;传感器是工艺设备中的关键共性技术。该识别与分析结果可对纺织服装产业的技术创新有一定的促进作用。

关键词: 关键共性技术, LDA主题模型, 社会网络分析, 智能纺织

Abstract: "Research on key generic technologies serves as a crucial support for advancing the innovative development of China's manufacturing industry. Accurate identification of key generic technologies has become a significant impetus for accelerating the transformation and upgrading of the industry. After extensively reviewing relevant references, it is found that research on the identification of key generic technologies has been applied in fields such as new materials, artificial intelligence, and new energy. However, no scholars have conducted relevant research in the textile and garment industry to date. Therefore, this paper focuses on the field of textile and garment, identifies the key generic technologies in this field, and provides substantive guidance and suggestions for relevant enterprises and departments. Firstly, the patent database of China National Intellectual Property Administration is used as the data source to collect patent data in the field of intelligent textiles. Based on the patent data, preprocessing operations are conducted. Subsequently, feature keyword extraction and the construction of a network relationship diagram are carried out to gain a preliminary understanding of the patent data, including the key classifications and correlation of patents in this industry. Secondly, the LDA topic model is used to uncover hidden high-intensity technical topics. Then, screening is conducted based on the ""co-occurrence"" and ""criticality"" of these technical topics. The co-occurrence degree of technical topics is evaluated through the co-occurrence rate index, so as to summarize the generic technologies. Afterwards, network analysis methods are applied, and the criticality of each node is quantified using three topological indicators: degree centrality, closeness centrality and structure holes. This further identifies the key generic technologies. The field of intelligent textiles represents the level of automation, informatization, intellectualization and digitalization of the entire industry, playing a significant role in promoting the digital transformation and high-quality development of the entire industry. Therefore, this paper focuses on the field of intelligent textiles within the textile and garment sector. The results indicate that graphene fibers, flexible nanofibers, polyurethane finishing agents, intelligent sewing technology, superhydrophobic coating technology, sensors, hydrogel conductive technology, carbon fiber composites, grafting technology, and 3D printing technology are the key generic technologies in this field. Finally, according to the identification results and the current development trends of the industry, substantive guidance and suggestions are provided for relevant enterprises and governments. It is recommended that related enterprises choose sustainable materials, manage production processes environmentally friendly, and strengthen technological innovation. For relevant governments, it is suggested to actively promote the concept of circular economy to relevant enterprises and consumers, and issue strong incentive policies to encourage enterprises to strive for upward development."

Key words: key generic technologies, LDA topic model, social network analysis, intelligent textiles

中图分类号: