A Method for Reusing Wind Turbine Assembly Processes Integrating Geometric and Process Semantics

doi:10.16183/j.cnki.jsjtu.2024.166

Abstract

Abstract:

To address the complexity of the knowledge system amassed during the historical assembly of wind turbines, which poses challenges to processing and leveraging this knowledge for designing new assembly processes, a method that integrates geometric and process semantics for reusing wind turbine assembly processes is developed. First, an assembly process information model based on knowledge graph-based wind turbine (KG-WT) is proposed, providing a unified and standardized representation of geometric and process semantics in the assembly process. Then, an extraction method for geometric semantics from glTF models and process semantics from assembly process documents is introduced, which extracts geometric and process semantic elements and maps them to entities in the knowledge graph data layer. Finally, a graph matching neural network is employed to match the assembly body or unit to be designed with the historical assembly knowledge base, thereby retrieving reusable similar assembly process items. The historical assembly knowledge is effectively reused in the design of the internal ring gear and connector assembly processes for wind turbine nacelles. The results indicate that the method proposed can provide a theoretical foundation for enhancing the efficiency and accuracy of assembly process design.

Key words: knowledge graph, assembly process information modelling, knowledge extraction, assembly process reuse, wind turbine

CLC Number:

TP183

LIU Mingfei, LING Wei, WANG Sen, BAO Jinsong. A Method for Reusing Wind Turbine Assembly Processes Integrating Geometric and Process Semantics[J]. Journal of Shanghai Jiao Tong University, 2026, 60(5): 809-823.

Figures/Tables 21

Fig.1

Tab.1

Fig.2

Tab.2

Fig.3

Fig.4

Fig.5

Fig.6

Fig.7

Tab.3

Fig.8

Fig.9

Fig.10

Fig.11

Tab.4

Fig.12

Fig.13

Fig.14

Tab.5

Tab.6

Fig.15

References 24

[1]	李耀华, 孔力. 发展太阳能和风能发电技术加速推进我国能源转型[J]. 中国科学院院刊, 2019, 34(4): 426-433.
	LI Yaohua, KONG Li. Developing solar and wind power generation technology to accelerate China’s energy transformation[J]. Bulletin of Chinese Academy of Sciences, 2019, 34(4): 426-433.
[2]	芮晓明, 尹文良. 差动调速的风电机组传动特性研究[J]. 中国机械工程, 2019, 30(9): 1034-1040.
	RUI Xiaoming, YIN Wenliang. Research on transmission performances of wind turbines with differential speed regulation[J]. China Mechanical Engineering, 2019, 30(9): 1034-1040.
[3]	GONG H Q, SHI L L, ZHAI X, et al. Assembly process case matching based on a multilevel assembly ontology method[J]. Assembly Automation, 2022, 42(1): 80-98. doi: 10.1108/AA-05-2021-0065 URL
[4]	AVDEENKO T V, MAKAROVA E S. Knowledge representation model based on case-based reasoning and the domain ontology: Application to the IT consultation[J]. IFAC-PapersOnLine, 2018, 51(11): 1218-1223. doi: 10.1016/j.ifacol.2018.08.424 URL
[5]	宁伟航, 刘检华, 熊辉, 等. 基于案例推理和词移距离的复杂产品装配现场技术问题管控方法[J]. 计算机集成制造系统, 2022, 28(3): 663-675.
	NING Weihang, LIU Jianhua, XIONG Hui, et al. On-site technical problem control method of complex product assembly process using case-based reasoning and word mover’s distance[J]. Computer Integrated Manufacturing Systems, 2022, 28(3): 663-675.
[6]	严键, 陈友玲, 刘文科. 基于本体映射的零件工艺实例重用方法研究[J]. 计算机应用研究, 2012, 29(1): 177-180.
	YAN Jian, CHEN Youling, LIU Wenke. Research on approach of part process case reusing based on ontology mapping[J]. Application Research of Compu-ters, 2012, 29(1): 177-180.
[7]	QIAO L H, QIE Y F, ZHU Z W, et al. An ontology-based modelling and reasoning framework for assembly sequence planning[J]. The International Journal of Advanced Manufacturing Technology, 2018, 94(9): 4187-4197. doi: 10.1007/s00170-017-1077-4 URL
[8]	HE Y, HAO C P, WANG Y L, et al. An ontology-based method of knowledge modelling for remanufacturing process planning[J]. Journal of Cleaner Production, 2020, 258: 120952. doi: 10.1016/j.jclepro.2020.120952 URL
[9]	MABKHOT M M, AL-SAMHAN A M, HIDRI L. An ontology-enabled case-based reasoning decision support system for manufacturing process selection[J]. Advances in Materials Science and Engineering, 2019, 2019: 2505183.
[10]	顾星海. 基于知识图谱的复杂产品装配工艺智能设计方法[D]. 上海: 东华大学, 2022.
	GU Xinghai. Intelligent design method of complex product assembly process based on knowledge map[D]. Shanghai: Donghua University, 2022.
[11]	王裴岩, 张桂平, 翟顺龙, 等. 基于多核学习的装配工艺过程重用[J]. 计算机集成制造系统, 2018, 24(7): 1850-1857.
	WANG Peiyan, ZHANG Guiping, ZHAI Shunlong, et al. Multiple kernel learning based assembly process reuse[J]. Computer Integrated Manufacturing Systems, 2018, 24(7): 1850-1857.
[12]	陈俊皓, 贾晓亮. 基于实例学习的飞机装配序列智能规划方法[J]. 航空制造技术, 2022, 62(12): 77-85.
	CHEN Junhao, JIA Xiaoliang. An approach for aircraft assembly sequence intelligent planning based on case learning[J]. Aeronautical Manufacturing Technology, 2022, 62(12): 77-85.
[13]	曹勇. 基于数据挖掘的工艺知识发现与重用研究[D]. 济南: 山东大学, 2019.
	CAO Yong. Research on process knowledge discovery and reuse based on data mining[D]. Jinan: Shandong University, 2019.
[14]	卫慧, 陈鹏, 张芮菡, 等. 基于长短期记忆网络的大型漂浮式风力发电机平台运动极短期预报方法[J]. 上海交通大学学报, 2023, 57(Sup.1): 37-45. doi: 10.16183/j.cnki.jsjtu.2023.S1.27
	WEI Hui, CHEN Peng, ZHANG Ruihan, et al. Extremely short-term forecasting method for large-scale floating wind turbine platform motion based on long and short-term memory networks[J]. Journal of Shanghai Jiao Tong University, 2023, 57(Sup.1): 37-45.
[15]	郭加伟, 许志杰, 何其昌. 虚拟装配手势库设计及人机工效评估优化[J]. 上海交通大学学报, 2022, 56(2): 127-133. doi: 10.16183/j.cnki.jsjtu.2021.095
	GUO Jiawei, XU Zhijie, HE Qichang. Design of a virtual assembly gesture library and optimization of ergonomics evaluation[J]. Journal of Shanghai Jiao Tong University, 2022, 56(2): 127-133.
[16]	花豹, 周彬, 顾星海, 等. 基于图神经网络的工艺表格语义相似性度量[J]. 计算机集成制造系统, 2022, 28(12): 3803-3819.
	HUA Bao, ZHOU Bin, GU Xinghai, et al. Semantic similarity measurement of process table based on graph neural network[J]. Computer Integrated Manufacturing Systems, 2022, 28(12): 3803-3819.
[17]	谢腾, 杨俊安, 刘辉. 基于BERT-BiLSTM-CRF模型的中文实体识别[J]. 计算机系统应用, 2020, 29(7): 48-55.
	XIE Teng, YANG Jun’an, LIU Hui. Chinese entity recognition based on BERT-BiLSTM-CRF model[J]. Computer Systems & Applications, 2020, 29(7): 48-55.
[18]	LI Y J, GU C J, DULLIEN T, et al. Graph mat-ching networks for learning the similarity of graph structured objects[DB/OL]. (2019-04-29) [2024-05-01]. http://arxiv.org/abs/1904.12787.
[19]	FAN C, LI Y. Research on Chinese word segmentation based on conditional random fields[M]// HUANGD S, JOK H, LIJ Q, et al. Lecture notes in computer science. Cham, Snitzerland: Springer, 2021: 316-326.
[20]	任智慧, 徐浩煜, 封松林, 等. 基于LSTM网络的序列标注中文分词法[J]. 计算机应用研究, 2017, 34(5): 1321-1324.
	REN Zhihui, XU Haoyu, FENG Songlin, et al. Sequence labeling Chinese word segmentation method based on LSTM networks[J]. Application Research of Computers, 2017, 34(5): 1321-1324.
[21]	QIN P D, XU W R, GUO J. A novel negative sampling based on TFIDF for learning word representation[J]. Neurocomputing, 2016, 177: 257-265. doi: 10.1016/j.neucom.2015.11.028 URL
[22]	YU D J, FANG A R, XU Z S. Topic research in fuzzy domain: Based on LDA topic modelling[J]. Information Sciences, 2023, 648: 119600. doi: 10.1016/j.ins.2023.119600 URL
[23]	KOTA K, MANJUNATHA A, SREE VIVEK S. CWE prediction using CVE description-the semantic similarity approach[J]. Procedia Computer Science, 2024, 235: 1167-1178. doi: 10.1016/j.procs.2024.04.111 URL
[24]	GROS D, HABERMANN T, KIRSTEIN G, et al. Anaphora resolution[J]. International Journal of Information Retrieval Research, 2018, 8(3): 33-45. doi: 10.4018/IJIRR URL

实体/关系类型	实体/关系类别及符号	实体/关系定义
装配单元实体,A_{E_g}	装配体总成单元,M_{E_g}	代表装配任务的目标,组件、零件以及连接件的总成
	组件单元,Z_{E_g}	代表相同系列的零部件
	零件单元,P_{E_g}	代表构成装配体产品的基本要素
	连接件单元,C_{E_g}	代表参与不同部件之间相互连接的单元
几何单元实体,G_{E_g}	点单元,V_{E_g}	代表构成几何体要素的顶点单元
	线单元,L_{E_g}	代表构成几何体表面的边界
	面单元,S_{E_g}	代表构成几何体的表面
装配元素关系,R_as	hasGroup	代表装配体与组件之间的广义包含关系
	hasPart	表示装配体、组件、零件之间的广义连接关系
	hasSub	代表零件单元间的广义连接关系
几何元素关系,R_ge	hasFace	代表零件表面由微观几何面构成
	hasEdge	代表几何表面的边界由边构成
	hasAnchors	代表边的两端由顶点构成
装配约束关系,R_ac	coincidentWith	代表装配单元间重合
	coaxialWith	代表装配单元间同轴
	connectedBy	代表零件在装配过程中连接关系
	perpendicularlWith	代表装配单元间垂直
	parellelWith	代表装配单元间平行
	diastanceWith	代表装配单元间存在固定间距

实体/关系类型	实体/关系类别及符号	实体/关系定义
装配工序类实体,A_{E_p}	装配工序类,P_{E_p}	代表装配体完成组装的某个步骤
	工序间类,W_{E_p}	代表装配件在前后工序间的装配状态
	工装夹具类,F_{E_p}	代表构成装配体产品的基本要素
装配工步类实体,G_{E_p}	工步类,M_{E_p}	代表某一装配元素的配合过程
	工步间类,R_{E_p}	代表装配件在前后工步的引用
	零件类,Y_{E_p}	代表参与装配工步的装配件
	辅助工具类,T_{E_p}	代表在装配工步中起辅助作用的工具
	装配要求类,C_{E_p}	代表工艺规范以及装配精度
	装配特征类,D_{E_p}	代表参与装配的装配件之间的几何特征
装配过程关系	before,after	代表先后执行装配动作
	with,stepTo	代表装配任务、动作执行顺序
	cycle	代表循环执行装配动作
	hasTool	代表操作过程中需要(has)用到某种工具
	hasAF	代表参与装配零件的装配特征
	executedTo	代表装配操作的执行对象
	priorTo	代表装配过程中的优先级关系
	matingWith	代表装配零件之间的配合关系

组合类别	图谱关系
装配零件P_{E_p}/辅助工具T_{E_p},装配零件P_{E_p}/辅助工具T_{E_p}+介宾	before/after/with
装配工步M_{E_p}/装配零件P_{E_p},辅助工具T_{E_p}+主谓/间宾/前宾	hasTool
装配工步M_{E_p},装配零件P_{E_p}+主谓/间宾/前宾	executedTo
装配工序P_{E_p},装配工序P_{E_p}	stepTo
装配工序P_{E_p},装配工步M_{E_p}/装配工步M_{E_p},装配工步M_{E_p}	with/priorTo
装配零件P_{E_p}/装配特征D_{E_p}	hasAF
装配零件P_{E_p}/装配零件P_{E_p}	connectedBy
v.:准备/预备/备用/筹备	prepare
上/顶部+介宾,下/底部+介宾,前+介宾,后+介宾	on/under/before/after

模型	P	R	F₁
CRF++^[19]	0.925	0.916	0.921
LSTM^[20]	0.957	0.952	0.955
BERT-BiLSTM-CRF	0.967	0.963	0.965

工步编号	操作内容	装配特征	装配动作	装配资源	装配要求
1-1	清洗	平面-119(面)		防锈型金属清洗剂
1-2	定位	圆柱面-122(面)		定位销专用工装
1-3	安装	平面-119(面)、平面-87(面)	吊装	吊具吊装	定位销孔应对齐保证联接体没有装反
1-4	连接		拧	棘轮扳手	螺栓需要润滑
1-5	检测	平面-119(面)、平面-87(面)		塞尺	间隙为0.45~0.75 mm