基于知识图谱的空管特情处置决策支持方法

doi:10.12305/j.issn.1001-506X.2024.12.20

摘要/Abstract

摘要：

空中交通管制特情是民航运输过程中遭遇的紧急特殊情况, 对其的处置原则是尽可能的精准和高效。若未能有效处置, 将引发重大飞行事故。然而, 传统的特情处置依赖于人工, 难以满足精准和高效的要求。利用知识图谱技术对空中交通管制特情信息进行知识抽取、表示和管理, 并用于辅助空中交通管制人员, 方便其进行特情处置, 可有效提升空中交通管制特情应急处置效率。因此, 提出一种自顶向下的空中交通管制特情处置知识图谱构建方法。首先, 自顶向下定义知识图谱的概念、关系及其知识架构, 形成模式层。接着, 考虑到空中交通管制特情案例记录文本训练量较小且领域性实体较多的特点, 采用融合双向长短时记忆(bi-directional long short-term memory, BiLSTM)网络深度学习模型和规则知识的实体抽取模型双向转换编码器(bi-directional encoder representations from transformers, BERT)-BiLSTM-条件随机场(conditional radom fields, CRF)+正则表达式(regular expression, RE)抽取实体。在此基础上, 利用BiLSTM+自注意力(self-attention, SA)模型对实体间关系进行抽取。之后, 采用Jaccard相关系数进行知识融合。最后, 利用Neo4j图数据库对构建的空中交通管制特情处置知识图谱进行可视化, 并对其在民航空中交通管制特情处置决策支持中的应用前景进行分析, 为空中交通管制部门的实际应用提供参考。

关键词: 知识图谱, 特情处置, 空中交通管制, 深度学习, 决策支持

Abstract:

Air traffic control (ATC) special situation is a special emergency situation encountered in the process of civil aviation transportation, and the principle of ATC disposal is as accurate and efficient as possible. If it is not disposed effectively, it will lead to major flight accidents. However, the traditional special situation disposal relies on manual labor, which is difficult to meet the requirements of precision and efficiency. Knowledge graph technology is used to extract, represent and manage ATC special situation information, and to assist ATC personnel in special situation disposal, which can effectively improve the emergency disposal efficiency of ATC special situation. Therefore, a top-down knowledge graph construction method for ATC special situation disposal is proposed. Firstly, the concept, relationship and knowledge structure of knowledge graph are defined from top to bottom to form a pattern layer. Then, considering the small amount of training in ATC special situation case records and the large number of domain entities, the entity extraction model which combined bi-directional long short-term memory (BiLSTM) network deep learning model and rule knowledge entity extraction model of bi-directional encoder representations from transformers (BERT)-BiLSTM-conditional radom fields (CRF)+regular expression (RE) is adopted to extract entities. On this basis, the BiLSTM+self-attention (SA) model is used to extract the relationships among entities. After that, the Jaccard correlation coefficient is used for knowledge fusion. Finally, the knowledge graph of the constructed ATC special situation disposal is visualized by using Neo4j graph database, and its application prospect in decision support of ATC special situation disposal in civil aviation is analyzed, providing references for the practical application of ATC department.

Key words: knowledge graph, special situation disposal, air traffic control (ATC), deep learning, decision support

中图分类号:

V355.1+1

彭珂, 王华伟, 侯召国, 曾啸寒, 罗通. 基于知识图谱的空管特情处置决策支持方法[J]. 系统工程与电子技术, 2024, 46(12): 4116-4127.

Ke PENG, Huawei WANG, Zhaoguo HOU, Xiaohan ZENG, Tong LUO. Decision support method of ATC special situation disposal based on knowledge graph[J]. Systems Engineering and Electronics, 2024, 46(12): 4116-4127.

图/表 18

表1

图1

图2

图3

图4

图5

图6

表2

图7

图8

表3

表4

表5

图9

表6

图10

图11

图12

参考文献 30

1	彭珂, 王华伟, 倪晓梅, 等. 基于TextRank的空管特情案例特征提取技术[J]. 航空计算技术, 2023, 53 (4): 56- 60.
	PENG K , WANG H W , NI X M , et al. Feature extraction technology of air traffic control case based on TextRank[J]. Aeronautical Computing Technique, 2023, 53 (4): 56- 60.
2	汪磊, 王朔, 高杉, 等. 面向特情的航线飞行员视觉注意分配研究[J]. 中国安全科学学报, 2023, 33 (1): 214- 220.
	WANG L , WANG S , GAO S , et al. A study on airline pilots' visual attention allocation during emergency[J]. China Safety Science Journal, 2023, 33 (1): 214- 220.
3	LEBEDEV V G , KULIDA E L . Advanced technologies for supporting operational decision-making in civil aviation[J]. Advances in Systems Science and Applications, 2021, 21 (3): 12- 21.
4	叶欣智, 尚磊, 董旭柱, 等. 面向配电网故障处置的知识图谱研究与应用[J]. 电网技术, 2022, 46 (10): 3739- 3749.
	YE X Z , SHANG L , DONG X Z , et al. Knowledge graph for distribution network fault handling[J]. Power System Technology, 2022, 46 (10): 3739- 3749.
5	郭榕, 杨群, 刘绍翰, 等. 电网故障处置知识图谱构建研究与应用[J]. 电网技术, 2021, 45 (6): 2092- 2100.
	GUO R , YANG Q , LIU S H , et al. Construction and application of power grid fault handing knowledge graph[J]. Power System Technology, 2021, 45 (6): 2092- 2100.
6	AMIT S. Introducing the knowledge graph: things, not strings[EB/OL]. [2023-05-12]. https://www.blog.google/products/search/introducing-knowledgegraphthings not.
7	王得强, 吴军, 关立文. 结合知识图谱的行业知识库构建方法研究[J]. 制造技术与机床, 2022, (8): 74- 80.
31	LEE K, HE L H, ZETTLEMOYER L. Higher-order corefe-rence resolution with coarse-to-fine inference[C]//Proc. of the North American Chapter of the Association for Computational Linguistics, 2018: 687-692.
7	WANG D Q , WU J , GUAN L W . Industry knowledge base construction based on knowledge graph[J]. Manufacturing Technology & Machine Tool, 2022, (8): 74- 80.
8	王萌, 王昊奋, 李博涵, 等. 新一代知识图谱关键技术综述[J]. 计算机研究与发展, 2022, 59 (9): 1947- 1965.
	WANG M , WANG H F , LI B H , et al. Survey on key techno-logies of new generation knowledge graph[J]. Journal of Computer Research and Development, 2022, 59 (9): 1947- 1965.
9	XIONG S H , CHEN Z S , CHANG J P , et al. On extended power average operators for decision-making: a case study in emergency response plan selection of civil aviation[J]. Compu-ters Industrial Engineering, 2019, 130, 258- 271.
10	WANG X , LIANG X D , LI X Y , et al. An integrated multiattribute group decision-making approach for risk assessment in aviation emergency rescue[J]. International Journal of Aerospace Engineering, 2022, 2022, 9713921.
11	SHEN W H , ZHONG Y F , WANG J J , et al. Construction and application of flood disaster knowledge graph based on multi-modal data[J]. Geomatics and Information Science of Wuhan University, 2023, 48 (12): 2009- 2018.
12	FANG W L , MA L , LOVE P E D , et al. Knowledge graph for identifying hazards on construction sites: integrating computer vision with ontology[J]. Automation in Construction, 2020, 119, 103310.
13	LIU X M , LU H K , LI H R . Intelligent generation method of emergency plan for hydraulic engineering based on knowledge graph-take the south-to-north water diversion project as an example[J]. LHB Hydroscience Journal, 2022, 108 (1): 2153629.
14	朱广宇, 张萌, 裔扬. 基于知识图谱的城市轨道交通突发事件演化结果预测[J]. 电子与信息学报, 2023, 45 (3): 949- 957.
	ZHU G Y , ZHANG M , YI Y . Prediction of evolution results of urban rail transit emergencies based on knowledge graph[J]. Journal of Electronics & Information Technology, 2023, 45 (3): 949- 957.
15	CHEN Y Y , ZHANG L B , HU J Q , et al. An emergency task recommendation model of long-distance oil and gas pipeline based on knowledge graph convolution network[J]. Process Safety and Environmental Protection, 2022, 2022 (167): 651- 661.
16	曹明, 王鹏, 左洪福, 等. 民用航空发动机故障诊断与健康管理现状、挑战与机遇Ⅱ: 地面综合诊断、寿命管理和智能维护维修决策[J]. 航空学报, 2022, 43 (9): 42- 81.
	CAO M , WANG P , ZUO H F , et al. Current status, challenges and opportunities of civil aero-engine diagnostics & health management Ⅱ: comprehensive off-board diagnosis, life management and intelligent condition based MRO[J]. Acta Aeronautica et Astronautica Sinica, 2022, 43 (9): 42- 81.
17	聂同攀, 曾继炎, 程玉杰, 等. 面向飞机电源系统故障诊断的知识图谱构建技术及应用[J]. 航空学报, 2022, 43 (8): 46- 62.
	NIE T P , ZENG J Y , CHENG Y J , et al. Knowledge graph construction technology and its application in aircraft power system fault diagnosis[J]. Acta Aeronautica et Astronautica Sinica, 2022, 43 (8): 46- 62.
18	吴闯, 张亮, 唐希浪, 等. 航空发动机润滑系统故障知识图谱构建及应用[J]. 北京航空航天大学学报, 2024, 50 (4): 1336- 1346.
	WU C , ZHANG L , TANG X L , et al. Construction and application of fault knowledge graph for aero-engine lubrication system[J]. Journal of Beijing University of Aeronautics and Astronautics, 2024, 50 (4): 1336- 1346.
19	王红, 李晗, 李浩飞. 民航突发事件领域本体关系提取方法的研究[J]. 计算机科学与探索, 2020, 14 (2): 285- 293.
	WANG H , LI H , LI H F . Research of relation extraction method of civil aviation emergency domain ontology[J]. Journal of Frontiers of Computer Science and Technology, 2020, 14 (2): 285- 293.
20	DEVLIN J, CHANG M W, LEE K, et al. BERT: pre-training of deep bidirectional transformers for language understanding[EB/OL]. [2023-05-13]. https://doi.org/10.48550/arXiv.1810.04805.
21	李思珍. 基于本体的行业知识图谱构建技术的研究与实现[D]. 北京: 北京邮电大学, 2019.
	LI S Z. The research and implementation of ontology-based entriprise knowledge graph construction[D]. Beijing: Beijing University of Posts and Telecommunications, 2019.
22	张逸, 李渴, 邵振国, 等. 基于标准文件的电能质量领域本体构建方法[J]. 电力系统自动化, 2020, 44 (17): 102- 110.
	ZHANG Y , LI K , SHAO Z G , et al. Standard files based ontology construction method in power quality domain[J]. Automation of Electric Power Systems, 2020, 44 (17): 102- 110.
23	段文昱, 朱继召, 赵浩楠, 等. 面向武器装备领域的实体及关系抽取方法研究[J]. 中国电子科学研究院学报, 2022, 17 (12): 1165- 1172.
	DUAN W Y , ZHU J Z , ZHAO H N , et al. Research on entity and relation extraction for weapons and equipment field[J]. Journal of China Academg of Electronics and Information Technology, 2022, 17 (12): 1165- 1172.
24	李冬梅, 张扬, 李东远, 等. 实体关系抽取方法研究综述[J]. 计算机研究与发展, 2020, 57 (7): 1424- 1448.
	LI D M , ZHANG Y , LI D Y , et al. Review of entity relation extraction methods[J]. Journal of Computer Research and Development, 2020, 57 (7): 1424- 1448.
25	罗欣, 陈艳阳, 耿昊天, 等. 基于深度强化学习的文本实体关系抽取方法[J]. 电子科技大学学报, 2022, 51 (1): 91- 99.
	LUO X , CHEN Y Y , GENG H T , et al. Entity relationship extraction from text data based on deep reinforcement learning[J]. Journal of University of Electronic Science and Technology of China, 2022, 51 (1): 91- 99.
26	胡杰, 李源洁, 耿號, 等. 基于深度学习的汽车故障知识图谱构建[J]. 汽车工程, 2023, 45 (1): 52-60, 85.
	HU J , LI Y J , GENG H , et al. Construction of vehicle fault knowledge graph based on deep learning[J]. Automotive Engineering, 2023, 45 (1): 52-60, 85.
27	COHEN W W, RICHMAN J. Learning to match and cluster large high-dimensional data sets for data integration[C]//Proc. of the 8th ACM Sigkdd International Conference on Knowledge Discovery and Data Mining, 2002: 475-480.
28	CLARK K, MANNING C D. Deep reinforcement learning for mention-ranking coreference models[C]//Proc. of the Confe-rence on Empirical Methods in Natural Language Processing, 2016: 2256-2262.
29	LEE K, HE I H, LEWIS M. et al. End-to-end neural corefe-rence resolution[C]//Proc. of the Conference on Empirical Methods in Natural Language Processing, 2017: 188-197.
30	PETERS M E, NEUMANN M, LYYER M, et al. Dee contex- tualized word representations[C]//Proc. of the North American Chapter of the Association for Computational Linguistics, 2018: 2227-2237.

数据名称	数据类型	选用部分
规章	半结构化	空管一般规定
检查单	半结构化	人员、处置措施、涉及部门等
特情案例记录	非结构化	特情名称、特情原因、人员、设备、指令、设备指令、飞行状态等

模型参数	模型参数取值
批尺寸	64
随机失活率	0.5
训练轮次	50
学习率	0.000 3
优化器	Adam
BERT读取的最大序列长度	128

实体抽取模型	P	R	F₁值
BiLSTM	0.754 3	0.737 7	0.745 9
BiLSTM-CRF	0.842 8	0.842 5	0.822 7
BERT-BiLSTM-CRF	0.895 8	0.903 1	0.899 4
BERT-BiLSTM-CRF+RE	0.920 1	0.952 1	0.935 8

模型参数	模型参数取值
嵌入方式	100维字向量
随机失活率	0.5
BiLSTM隐藏层单元数	200
训练轮次	100
学习率	0.000 5

序号	语料整理结果
1	自动化系统短期冲突预警出现自动化系统出现短期冲突预警, 两机水平间隔29.6 km
2	自动化系统水平间隔监测自动化系统出现短期冲突预警, 两机水平间隔29.6 km
3	水平间隔短期冲突预警引起自动化系统出现短期冲突预警, 两机水平间隔29.6 km