基于KG-KGCN的某型动力装置风险分析方法

陈国兵; 曾国庆; 王越; 王学峰; 谢旭阳; 杨自春

doi:10.13224/j.cnki.jasp.20220007

基于KG-KGCN的某型动力装置风险分析方法

doi: 10.13224/j.cnki.jasp.20220007

海军工程大学动力工程学院，武汉 430033

基金项目: 国家自然科学基金（51609251）

详细信息

作者简介:
陈国兵（1982-），男，教授、博士生导师，博士，主要从事舰船动力及热力设备的科学管理研究。E-mail：chenguob@163.com

通讯作者:
杨自春（1967-），男，教授、博士生导师，博士，主要从事舰船动力及热力设备的科学管理研究。E-mail：yangzichun11@sina.com

中图分类号: V240.2；TP391
计量
- 文章访问数: 525
- HTML浏览量: 340
- PDF量: 54
- 被引次数: 0
出版历程
- 收稿日期: 2022-01-05
- 网络出版日期: 2023-06-09

Risk analysis method of a certain type of power plant based on KG-KGCN

College of Power Engineering，Naval University of Engineering，Wuhan 430033，China

摘要

摘要:
针对动力装置危险因素众多、风险数据繁杂且利用率不高等问题，以故障模式、影响及危害性分析等数据为基础，提出基于知识图谱和图神经网络的系统风险分析方法及算法，利用事件的知识完成链接预测，推理出引发事件的原因及影响等，以实现其精准的统计分析。本文以某型发动机为对象，对其进行风险分析和推理。结果表明：该方法有效解决了传统方法存在的表单繁琐、分析单一等问题，结果更加准确和全面，可使分析效率提升60%以上；并可及时开展风险推理和预测，实现从事后分析向主动预防的转变，为动力装置的智能运维提供了有力支持。
- 风险分析 /
- 知识图谱 /
- 知识图卷积网络 /
- 动力装置 /
- 故障模式、影响及危害性分析(FMECA)
Abstract:
Targeting the problems of numerous risk factors, complex risk data and low utilization rate of power plant, and based on the data of failure mode, impact and criticality analysis, a method and algorithm of system risk analysis was proposed using knowledge graph and graph neural network, and the knowledge of events were utilized to complete link prediction and deduce the cause and impact of events, so as to achieve accurate statistical analysis. A certain type of engine was taken as the object to carry out risk analysis and reasoning. Results showed that this method effectively solved the problems of the traditional method, such as cumbersome forms and single analysis, the results were more accurate and comprehensive, and the analysis efficiency can be improved by more than 60%; and it can carry out risk reasoning and prediction in time, and realize the transformation from post analysis to active prevention, providing a strong support for the intelligent operation and maintenance of power plants.
- risk analysis /
- knowledge graph /
- knowledge graph convolutional network /
- power plant /
- failure mode, effects and criticality analysis (FMECA)

HTML

图 1 系统风险知识图谱构建

Figure 1. Process of building knowledge graph of system risk

下载: 全尺寸图片幻灯片

图 2 实体及其关系结构图

Figure 2. Entities and relationships structure diagram

下载: 全尺寸图片幻灯片

图 3 KGCN的知识图谱推理过程

Figure 3. Reasoning process of knowledge graph of KGCN

下载: 全尺寸图片幻灯片

图 4 知识图谱实体两层感受野^[14]

Figure 4. Two-layer of receptive fields of the entity in the knowledge graph^[14]

下载: 全尺寸图片幻灯片

图 5 系统风险分析知识图谱统计与查询流程

Figure 5. Process of system risk analysis knowledge graph statistical and query

下载: 全尺寸图片幻灯片

图 6 基于KGCN的推理过程

Figure 6. Reasoning process based on KGCN

下载: 全尺寸图片幻灯片

图 7 KG-KGCN系统风险分析

Figure 7. Model of KG-KGCN system risk analysis

下载: 全尺寸图片幻灯片

图 8 发动机FMECA层次约定

Figure 8. Conventional level of engine FMECA

下载: 全尺寸图片幻灯片

图 9 发动机FMECA整体知识图谱

Figure 9. Overall knowledge graph of engine FMECA

下载: 全尺寸图片幻灯片

图 10 知识图谱结构模型

Figure 10. Structure model of knowledge graph

下载: 全尺寸图片幻灯片

图 11 下传动箱故障零件统计图谱

Figure 11. Faulty parts statistics graph in the lower gearbox

下载: 全尺寸图片幻灯片

图 12 分系统设备统计图谱

Figure 12. Equipment statistics graph of the part system

下载: 全尺寸图片幻灯片

图 13 子系统与模块之间查询界面

Figure 13. Interface of knowledge graph query between subsystems and modules

下载: 全尺寸图片幻灯片

图 14 低压压气机转子具体故障查询结果界面

Figure 14. Interface of query results of specific faults of the low-pressure compressor rotor

下载: 全尺寸图片幻灯片

图 15 放气阀漏气事件推理结果

Figure 15. Inference result of the air release valve leakage event

下载: 全尺寸图片幻灯片

图 16 高压压气机静子放气阀漏气故障树

Figure 16. FTA of the air release valve leakage event

下载: 全尺寸图片幻灯片

表 1 不同类型实体及其关系

Table 1. Different types of entities and their relationships

模型数据	数据结构	作用	对象范围
实体	标签	可被标记多个类别标签	系统、故障模式、故障原因、故障影响
实体	属性	可存储多个相关属性	编号、工况、严酷度等
关系	类型	所属关系类型	导致关系、包含关系
关系	属性	存储关系属性	关系名称、关系ID

下载: 导出CSV

参考文献(21)

[1]	BAHREBAR S,BLAABJERG F,WANG Huai,et al. A novel type-2 fuzzy logic for improved risk analysis of proton exchange membrane fuel cells in marine power systems application[J]. Energies,2018,11(4): 721. doi: 10.3390/en11040721
[2]	MORENO V C,COZZANI V. Integrated hazard identification within the risk management of industrial biological processes[J]. Safety Science,2018,103: 340-351. doi: 10.1016/j.ssci.2017.12.004
[3]	DUNJÓ J,FTHENAKIS V,VÍLCHEZ J A,et al. Hazard and operability (HAZOP) analysis. A literature review[J]. Journal of Hazardous Materials,2010,173(1/2/3): 19-32.
[4]	徐立,邓儒超,徐楚,等. 船舶核动力装置二回路故障分析及安全性研究[J]. 哈尔滨工程大学学报,2015,36(1): 130-133. XU Li,DENG Ruchao,XU Chu,et al. Failure analysis and safety studies of the second loop system of civil marine nuclear power plant[J]. Journal of Harbin Engineering University,2015,36(1): 130-133. (in Chinese)
[5]	屠秋野,张伟,蔡元虎,等. 民用航空动力技术风险评估的量化方法[J]. 航空动力学报,2010,25(11): 2403-2408. TU Qiuye,ZHANG Wei,CAI Yuanhu,et al. Quantity method of technical risk assessment for civil aero-engine[J]. Journal of Aerospace Power,2010,25(11): 2403-2408. (in Chinese)
[6]	AMIT S. Introducing the knowledge graph: things, not strings[EB/OL]. [2022-02-15].https: //www.blog.google/products/se-arch/introducing-knowledge-graph-things-not.
[7]	KRIZHANOVSKY A A,SMIRNOV A V. An approach to automated construction of a general-purpose lexical ontology based on Wiktionary[J]. Journal of Computer and Systems Sciences International,2013,52(2): 215-225. doi: 10.1134/S1064230713020068
[8]	魏卉子. 煤矿安全融合知识图谱构建研究[D]. 江苏徐州: 中国矿业大学, 2020. WEI Huizi. Study on the construction of coal mine safety integration knowledge map[D]. Xuzhou Jiangsu: China University of Mining and Technology, 2020. (in Chinese)
[9]	IYYER M, BOYD-GRABER J, CLAUDINO L, et al. A neural network for factoid question answering over paragraphs[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP). Stroudsburg, US: Association for Computational Linguistics, 2014: 633-644.
[10]	邱凌,张安思,李少波,等. 航空制造知识图谱构建研究综述[J]. 计算机应用研究,2022,39(4): 968-977. QIU Ling,ZHANG Ansi,LI Shaobo,et al. Survey on building knowledge graphs for aerospace manufacturing[J]. Application Research of Computers,2022,39(4): 968-977. (in Chinese)
[11]	郭静. 基于知识图谱的航空安全事件关联分析方法研究[D]. 天津: 中国民航大学, 2020. GUO Jing. Research on the correlation analysis of aviation safety incidents based on knowledge graph[D]. Tianjin: Civil Aviation University of China, 2020. (in Chinese)
[12]	杜党党,贾晓亮,王卓. 基于故障彩色图谱的航空发动机故障模式识别[J]. 计算机集成制造系统,2018,24(9): 2297-2305. DU Dangdang,JIA Xiaoliang,WANG Zhuo. Fault model identification method for aero engines based on fault color picture[J]. Computer Integrated Manufacturing Systems,2018,24(9): 2297-2305. (in Chinese)
[13]	BRUNA J, ZAREMBA W, SZLAM A, et al. Spectral networks and locally connected networks on graphs[EB/OL]. [2021-12-08]. https://arxiv.org/abs/1312.6203
[14]	WANG Hongwei, ZHAO Miao, XIE Xing, et al. Knowledge graph convolutional networks for recommender systems[C]//Proceedings of the 19th World Wide Web Conference. New York: ACM, 2019: 3307-3313.
[15]	ZHANG Ya,LU Mingming,LI Haifeng. Urban traffic flow forecast based on FastGCRNN[J]. Journal of Advanced Transportation,2020,2020: 1-9.
[16]	KAPOOR A, BEN Xue, LIU Luyang, et al. Examining COVID-19 forecasting using spatio-temporal graph neural networks[EB/OL]. [2021-12-15]. https://arxiv.org/abs/2007.03113
[17]	刘欢,李晓戈,胡立坤,等. 基于知识图谱驱动的图神经网络推荐模型[J]. 计算机应用,2021,41(7): 1865-1870. LIU Huan,LI Xiaoge,HU Likun,et al. Knowledge graph driven recommendation model of graph neural network[J]. Journal of Computer Applications,2021,41(7): 1865-1870. (in Chinese)
[18]	史运涛,党亚光,雷振伍,等. 基于图神经网络的社区户内燃气系统动态风险评估[J]. 安全与环境工程,2021,28(5): 1-9. SHI Yuntao,DANG Yaguang,LEI Zhenwu,et al. Dynamic risk assessment of community indoor gas system based on graph neural network[J]. Safety and Environmental Engineering,2021,28(5): 1-9. (in Chinese)
[19]	孟寒姣. 基于知识图谱的铁路电务事故智能问答系统的设计与实现[D]. 石家庄: 河北科技大学, 2020. MENG Hanjiao. Design and implementation of intelligent question answering system for railway electrical accidents based on knowledge graph[D]. Shijiazhuang: Hebei University of Science and Technology, 2020. (in Chinese)
[20]	何巍. 社交媒体数据可视化分析综述[J]. 科学技术与工程,2020,20(32): 13085-13090. HE Wei. An overview on social media visual analytics[J]. Science Technology and Engineering,2020,20(32): 13085-13090. (in Chinese)
[21]	王红,林海舟,卢林燕. 基于Att_GCN模型的知识图谱推理算法[J]. 计算机工程与应用,2020,56(9): 183-189. WANG Hong,LIN Haizhou,LU Linyan. Knowledge graph inference algorithm based on Att_GCN model[J]. Computer Engineering and Applications,2020,56(9): 183-189. (in Chinese)