中国机械工程 ›› 2025, Vol. 36 ›› Issue (06): 1206-1213.DOI: 10.3969/j.issn.1004-132X.2025.06.008

• 机械基础工程 • 上一篇    下一篇

风电设备情境知识图谱构建技术研究

石致远1,2;孔志伟2*;陈俊臻3;王淑营3   

  1. 1.西南交通大学机械工程学院,成都,610031
    2.东方电气集团科学技术研究院有限公司,成都,611731
    3.西南交通大学计算机与人工智能学院,成都,610031

  • 出版日期:2025-06-25 发布日期:2025-08-04
  • 作者简介:石致远,男,1989年生,高级工程师。研究方向为智能制造系统数字孪生应用技术。获省部级奖项2项,授权专利40余项。发表论文15篇。E-mail:shizy@dongfang.com。
  • 基金资助:
    四川省科技厅工业软件及信息安全重大科技专项(2022ZDZX0003)

Research on Construction Techniques for Wind Power Equipment Contextual Knowledge Graphs

SHI Zhiyuan1,2;KONG Zhiwei2*;CHEN Junzhen3;WANG Shuying3   

  1. 1.School of Mechanical Engineering,Southwest Jiaotong University,Chengdu,610031
    2.Intelligence Manufacturing Department,Dongfang Electric Academy of Science and Technology
    Co.,Ltd.,Chengdu,611731
    3.School of Computing and Artificial Intelligence,Southwest Jiaotong University,Chengdu,610031

  • Online:2025-06-25 Published:2025-08-04

摘要: 传统知识图谱构建方法未考虑知识的情境约束,难以有效表征风电等复杂机电设备海量知识间复杂的关联关系,限制了知识图谱在实际生产过程中的应用。提出了一种面向风电设备的情境知识图谱构建方法。首先抽取风电设备情境知识、模块元知识及项目定制产生的模块实例知识,结合形状约束语言(SHACL)构建了包含情境路径和属性值约束的本体模型,精准表征和抽取各类知识;然后提出了基于本体解析的情境知识子图可视化算法,通过解析本体中的情境知识类,为每类情境构建数据观测窗口,实现面向场景的知识子图多维可视化交互。实际应用结果表明,该方法能有效融合模块元知识与项目模块实例知识,满足风电设备知识的精准表征和多样化的应用场景需求。

关键词: 风电设备, 情境知识图谱, 情境语义约束, 可视化交互引擎, 形状约束语言

Abstract: Traditional methods for constructing knowledge graphs didnt consider the contextual constraints on knowledge, making it challenging to effectively represent the complex associative relationships among vast knowledge in complex electromechanical equipment like wind turbines. This limitation hindered the practical applications of knowledge graphs in the production processes. This paper proposed a method for constructing a context-aware knowledge graph tailored for wind turbine equipment. Initially, the method extracted contextual knowledge, module meta-knowledge, and module instance knowledge generated from project customization. Utilizing the SHACL, an ontology model was constructed, incorporating context paths and attribute value constraints, thereby precisely characterizing and extracting various knowledge types. Furthermore, an ontology parsing-based algorithm is introduced for visualizing contextual knowledge subgraphs. Through the parsing of contextual knowledge classes within the ontology, data observation windows were generated for each contextual class, facilitating the construction of multi-dimensional visualization interactions tailored to specific scenarios. Through practical applications, the proposed method effectively integrates module meta-knowledge with project-specific module instance knowledge, meeting the demands for precise representation and diverse application scenarios in wind turbine equipment knowledge.

Key words: wind power equipment, contextual knowledge graph, contextual semantic constraint, visualization interaction engine, shape constraint language(SHACL)

中图分类号: