HAZOP知识图谱构建方法

doi:10.16085/j.issn.1000-6613.2020-2004

摘要/Abstract

摘要：

危险与可操作分析方法（hazard and operability analysis，HAZOP）主要采用头脑风暴的形式，将讨论结果记录在纸质文档中，该方法过于依赖专家经验且大量现有的HAZOP信息并没有得到共享与复用。针对此问题，本文提出了结合自顶向下和自底向上的HAZOP知识图谱半自动构建方法。在解析HAZOP信息基础上，结合HAZOP分析国际标准IEC 61882，以危险事件角度设计了HAZOP本体规则，完成了模式层的构建；以现有HAZOP数据为基础，在BiLSTM-CRF命名实体识别模型基础上引入了Attention机制，实现了关键HAZOP信息的自动提取并利用图数据库进行数据存储，完成了数据层的构建。以某油品装置HAZOP分析报告为例构建HAZOP知识图谱，验证了构建方法的有效性。结果表明HAZOP知识图谱能够清晰地展示HAZOP知识之间的联系，快速提供现有HAZOP分析信息，辅助人工HAZOP的开展，降低人工成本与时间成本，也为后期项目的生产与维护提供知识支撑，实现了HAZOP信息的共享与复用。

关键词: 安全, 危险与可操作分析方法, 知识图谱, 模型, 命名实体识别, 神经网络

Abstract:

Hazard and operability analysis (HAZOP) takes the form of brainstorming, and it relies too much on expert experience. It is hard to share and reuse for the discussion results which were recorded in paper documents. Under the guidance of international standard IEC 61882, the ontology rules were developed through the perspective of dangerous events, and the model layer construction was completed after that. Then, based on the existing HAZOP analysis data, the attention mechanism was introduced on the BiLSTM-CRF model, which realized the automatic extraction of key HAZOP information and graph database was used for data storage to complete the construction of the data layer. Taking the HAZOP analysis report of oil equipment as an example, the HAZOP knowledge graph was constructed to verify the effectiveness of the construction method. It was proved that HAZOP knowledge graph can clearly demonstrate the relationship between HAZOP knowledge, quickly provide existing HAZOP information, assist the development of manual HAZOP, reduce labor and time costs, provide knowledge support for the production and maintenance of later projects, and realize the sharing and reuse of HAZOP information.

Key words: safety, hazard and operability analysis (HAZOP), knowledge graph, model, named entity recognition, neural networks

中图分类号:

TH391.7

李芳国, 张贝克, 高东. HAZOP知识图谱构建方法[J]. 化工进展, 2021, 40(8): 4666-4677.

LI Fangguo, ZHANG Beike, GAO Dong. Construction method of HAZOP knowledge graph[J]. Chemical Industry and Engineering Progress, 2021, 40(8): 4666-4677.

图/表 13

参考文献 27

1	吴重光. 危险与可操作性分析(HAZOP)基础及应用[M]. 北京: 中国石化出版社, 2012.
	WU Chongguang. Hazard and operability analysis basic and application[M]. Beijing: China Petrochemical Press, 2012.
2	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.
3	VIEGAS R A, FRANCISCO DE ASSIS DA SILVA MOTA, COSTA A P C S, et al. A multi-criteria-based hazard and operability analysis for process safety[J]. Process Safety and Environmental Protection, 2020, 144: 310-321.
4	SINGLE J I, SCHMIDT J, DENECKE J. Ontology-based support for hazard and operability studies[J]. International Journal of Safety and Security Engineering, 2020, 10(3): 311-319.
5	王峰, 高金吉, 张贝克, 等. 基于层次分析法的计算机辅助HAZOP分析技术[J]. 化工进展, 2008, 27(12): 2013-2018.
	WANG Feng, GAO Jinji, ZHANG Beike, et al. Computer aided HAZOP analysis technology based on AHP[J]. Chemical Industry and Engineering Progress, 2008, 27(12): 2013-2018.
6	SOWA J F. Principles of semantic networks: exploration in the representation of knowledge[M]. San Mateo, Calif: Morgan Kaufmann, 1991: 135-157.
7	PETER P S, HORROCKS I. OWL1.1 web ontology language overview[Z]. February, 2006, 63(45): 990-996.
8	SIDOR G. Introducing google mom-the only knowledge graph you’ll ever need[Z]. 2015.
9	李涓子, 侯磊. 知识图谱研究综述[J]. 山西大学学报(自然科学版), 2017, 40(3): 454-459.
	LI Juanzi, HOU Lei. Reviews on knowledge graph research[J]. Journal of Shanxi University (Natural Science Edition), 2017, 40(3): 454-459.
10	BIZER C, LEHMANN J, KOBILAROV G, et al. DBpedia—A crystallization point for the Web of Data[J]. Journal of Web Semantics, 2009, 7(3): 154-165.
11	XU B, XU Y, LIANG J Q, et al. CN-DBpedia: A never-ending Chinese knowledge extraction system[M]//Advances in Artificial Intelligence: From Theory to Practice. Cham: Springer International Publishing, 2017: 428-438.
12	HE L L, JIANG P Y. Manufacturing knowledge graph: a connectivism to answer production problems query with knowledge reuse[J]. IEEE Access, 2019, 7: 101231-101244.
13	YAN H H, YANG J, WAN J F. KnowIME: a system to construct a knowledge graph for intelligent manufacturing equipment[J]. IEEE Access, 2020, 8: 41805-41813.
14	REESE J T, UNNI D, CALLAHAN T J, et al. KG-COVID-19: a framework to produce customized knowledge graphs for COVID-19 response[J]. Patterns (New York, N Y), 2021, 2(1): 100155.
15	MAO S, ZHAO Y M, CHEN J H, et al. Development of process safety knowledge graph: a case study on delayed coking process[J]. Computers & Chemical Engineering, 2020, 143: 107094.
16	ZHENG X, WANG B, ZHAO Y M, et al. A knowledge graph method for hazardous chemical management: ontology design and entity identification[J]. Neurocomputing, 2021, 430: 104-111.
17	CATINO C A, UNGAR L H. Model-based approach to automated hazard identification of chemical plants[J]. AIChE Journal, 1995, 41(1): 97-109.
18	赵劲松, 赵利华, 崔林, 等. 基于案例推理的HAZOP分析自动化框架[J]. 化工学报, 2008, 59(1): 111-117.
	ZHAO Jinsong, ZHAO Lihua, CUI Lin, et al. Case based reasoning framework for automating HAZOP analysis[J]. Journal of Chemical Industry and Engineering (China), 2008, 59(1): 111-117.
19	吴重光, 许欣, 纳永良, 等. 基于知识本体的过程安全分析信息标准化[J]. 化工学报, 2012, 63(5): 1484-1491.
	WU Chongguang, XU Xin, NA Yongliang, et al. Standardized information for process hazard analysis based on ontology[J]. CIESC Journal, 2012, 63(5): 1484-1491.
20	BATRES R, WEST M, LEAL D, et al. An upper ontology based on ISO 15926[J]. Computers & Chemical Engineering, 2007, 31(5/6): 519-534.
21	高东, 肖遥, 张贝克, 等. 基于知识本体的HAZOP信息标准化框架[J]. 化工进展, 2020, 39(6): 2510-2518.
	GAO Dong, XIAO Yao, ZHANG Beike, et al. HAZOP information standardization framework based on knowledge ontology[J]. Chemical Industry and Engineering Progress, 2020, 39(6): 2510-2518.
22	GREFF K, SRIVASTAVA R K, KOUTNÍK J, et al. LSTM: a search space odyssey[J]. IEEE Transactions on Neural Networks and Learning Systems, 2017, 28(10): 2222-2232.
23	LIPTON Z C. A critical review of recurrent neural networks for sequence learning[EB/OL]. 2015.
24	HUANG Z H, XU W, YU K. Bidirectional LSTM-CRF models for sequence tagging[J]. Computer Science, 2015: arXiv: .
25	BAHDANAU D, CHO K, BENGIO Y. Neural machine translation by jointly learning to align and translate[EB/OL]. 2014.
26	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//31st Conference on Neural Information Processing System (NIPS 2017), LongBeach, CA, USA, 2017
27	JOHN L, ANDREW M, PEREIRA FERNANDO C N. Conditional random fields: probabilistic models for segmenting and labeling sequence data[C]//ICML'01: Proceedings of the Eighteenth International Conference on Machine Learning, 2001: 282-289.

实体类型	开始标记	后续标记
设备	B-EQU	I-EQU
物料	B-MAT	I-MAT
状态	B-STA	I-STA
非实体标记	O	O

实体类型	开始标记	后续标记
设备	B-EQU	I-EQU
物料	B-MAT	I-MAT
状态	B-STA	I-STA
非实体标记	O	O

模型	精确度（P）	召回率（R）	F₁
CRF	71.71	72.14	71.93
BiLSTM-CRF	75.54	78.97	77.22
BiLSTM-Self-Attention-CRF	80.58	83.00	81.77
BiLSTM-Multi-Head Attention-CRF	88.70	90.48	89.58

模型	精确度（P）	召回率（R）	F₁
CRF	71.71	72.14	71.93
BiLSTM-CRF	75.54	78.97	77.22
BiLSTM-Self-Attention-CRF	80.58	83.00	81.77
BiLSTM-Multi-Head Attention-CRF	88.70	90.48	89.58

序号	参数	引导词	后果	原因	防护措施	建议
1	R-5611101温度	过高	反应器超温、泄漏，现场发生火灾爆炸，人员中毒伤亡	1.除氧水流量过小 2.R-5611101液位过低 3.还原催化剂填装过量	1.TI0108-0113 AH 2.TI0201-0219 AH 3.PDI0202A/B AH 4.PDI0203A/B AH	1.需考虑增设PDI0202A/B和PDI0203A/B压差低报 2.建议设计单位增加其他防护措施降低温度超高的危害
2	R-5611101压力	过高	反应器损坏，密封点泄漏，现场发生火灾爆炸	1.原料来气压力过高 2.R-5611101反应异常	1.PY0201B SAHH 2.PY0201B联锁	1.考虑在新鲜气来气流程设置安全泄压设施，重新核算R-5611101安全泄放阀尺寸 2.建议设计单位在设计安全泄放阀时，考虑增加一组安全阀来提高安全可靠性，降低风险
3	R-5611101液位	过高	压缩机带液，设备损坏	1.重质蜡外送流程不畅	1.PDI0202A/B AH 2.PDI0201A/B AH 3.LY0601C SAHH 4.LY0601C联锁 5.LY0601B AH 6.PIS0201A/B/C AL
4		过低	反应器超温，设备损坏	1.渣蜡排放流程异常 2.程控阀门故障	1.PDI0202A/B AH 2.PDI0201A/B AH AHH 3.TI0108-0113 AH