基于船厂分段时空数据的分段状态识别及转运监测

doi:10.16183/j.cnki.jsjtu.2021.287

摘要/Abstract

摘要：

船舶分段转运保障了分段在各个工艺之间的有序流动,但消耗了大量成本.船厂管理者需要监测实际转运过程,特别是监测因堆场内分段互相阻挡和返工产生的两类非生产性转运.S船厂由于场地紧张必须一场多用,现有监测技术难以从分段时空数据中获取分段状态,进而难以实现两类非生产性转运的监测.针对这个问题,研究了转运过程中分段状态的时序转化规律和耗时特征,建立了4个隐马尔可夫模型并使用有监督的方法学习其参数,通过维特比算法实现了分段状态识别,其在测试集上的准确率最高达到了93.5%.将其中1个隐马尔可夫模型应用于船厂分段时空数据,实现了船厂的两类非生产性转运的监测,根据监测结果为优化船厂分段转运过程提出初步建议.

关键词: 船舶分段转运, 分段状态识别, 分段转运监测, 隐马尔可夫模型

Abstract:

Ship block transfer is important to the orderly flow of blocks between crafts, which is costly. Shipyard managers have to monitor the actual transfers, especially the unproductive transfers that occur when blocks are obstructed or reworked. A high-load shipyard, called S, often uses one site for multiple purposes, and the difficulties in obtaining the state of ship blocks through the time-site data of blocks provided by the existing monitoring technology make it difficult to monitor two types of unproductive transfers. To address this problem, four hidden Markov models whose parameters are calculated by a supervised approach are proposed, and a Viterbi algorithm based method is proposed to identify the state of blocks, achieving an accuracy of up to 93.5% on the test dataset. One of the hidden Markov models is applied to the time-site data of blocks to monitor two types of unproductive transfers in shipyards. Preliminary suggestions for improving the blocks transfer process based on monitoring results are proposed.

Key words: ship block transfer, block state identification, block transfer monitoring, hidden Markov model (HMM)

中图分类号:

TP391
U671

陈俊宇, 田凌. 基于船厂分段时空数据的分段状态识别及转运监测[J]. 上海交通大学学报, 2023, 57(1): 24-35.

CHEN Junyu, TIAN Ling. Ship Block State Identification and Transfer Monitoring Based on Time-Site Data of Blocks[J]. Journal of Shanghai Jiao Tong University, 2023, 57(1): 24-35.

图/表 14

表1

图1

图2

图3

表2

表3

图4

表4

表5

图5

表6

图6

图7

图8

参考文献 22

[1]	KIM K T, KIM H J, CHOI S H. Heuristic algorithms for assigning ship assembly blocks to storage locations at a shipyard[J]. Journal of International Logistics & Trade, 2017, 15(3): 83-90.
[2]	PARK C, SEO J. Assembly block storage location assignment problem: Revisited[J]. Production Planning & Control, 2009, 20(3): 216-226.
[3]	陶宁蓉, 蒋祖华, 刘建峰. 带时间窗约束的船体分段空间调度问题[J]. 计算机集成制造系统, 2010, 16(12): 2674-2679.
	TAO Ningrong, JIANG Zuhua, LIU Jianfeng. Spatial scheduling problem with time window constraint for block assembly in shipbuilding[J]. Computer Integrated Manufacturing Systems, 2010, 16(12): 2674-2679.
[4]	陶宁蓉. 船舶分段建造过程中的资源调度优化研究[D]. 上海: 上海交通大学, 2013.
	TAO Ningrong. Research on resource scheduling problems during ship block assembly process[D]. Shanghai: Shanghai Jiao Tong University, 2013.
[5]	陈凯, 蒋祖华, 刘建峰, 等. 带有进场时间窗的船舶分段堆场调度[J]. 上海交通大学学报, 2016, 50 (9): 1390-1398.
	CHEN K, JIANG ZH, LIU JF, et al. Shipbuilding yard scheduling with block inbound time window[J]. Journal of Shanghai Jiao Tong University, 2016, 50 (9): 1390-1398.
[6]	李柏鹤, 蒋祖华, 陶宁蓉, 等. 基于混合优化算法的船舶分段堆场间调度研究[J]. 哈尔滨工程大学学报, 2018, 39 (12): 2025-2032.
	LI Baihe, JIANG Zuhua, TAO Ningrong, et al. Research on dispatch of blocks between stockyards based on hybrid optimization algorithm[J]. Journal of Harbin Engineering University, 2018, 39(12): 2025-2032.
[7]	李柏鹤, 蒋祖华, 陶宁蓉, 等. 考虑平板车合作运输的船舶分段堆场间调度[J]. 上海交通大学学报, 2020, 54(7): 718-727.
	LI Baihe, JIANG Zuhua, TAO Ningrong, et al. Ship block transportation scheduling considering cooperative transportation of flatcars[J]. Journal of Shanghai Jiao Tong University, 2020, 54(7): 718-727.
[8]	MUN S, AN J, LEE J. Real time transporter locating system for shipyard through GNSS and IMU sensor[J]. Journal of the Society of Naval Architects of Korea, 2019, 56(5): 439-446. doi: 10.3744/SNAK.2019.56.5.439 URL
[9]	PARK J G, OH J S, LEE S R, et al. Location determination system for transport path optimization of block transporter[J]. The Journal of Korea Information and Communications Society, 2014, 39C(7): 589-596.
[10]	SONG K S, LEE S, CHO D Y. Smart device based localization for ship block logistics[J]. Journal of Korea Multimedia Society, 2012, 15(12): 1506-1516. doi: 10.9717/kmms.2012.15.12.1506 URL
[11]	CHA J H, CHO D Y, RUY W S, et al. Development of optimal planning system for operating transporters in shipyard[J]. Korean Journal of Computational Design and Engineering, 2016, 21(2): 177-185. doi: 10.7315/CADCAM.2016.177 URL
[12]	张恒, 向祖权, 王冲, 等. 船体分段外场物流实时综合监控系统[J]. 造船技术, 2019(4): 83-88.
	ZHANG Heng, XIANG Zuquan, WANG Chong, et al. Real-time integrated monitoring system of hull block outfield logistics[J]. Marine Technology, 2019(4): 83-88, 92.
[13]	LEE S K, KIM B, HUH M, et al. Mining transportation logs for understanding the after-assembly block manufacturing process in the shipbuilding industry[J]. Expert Systems With Applications, 2013, 40(1): 83-95. doi: 10.1016/j.eswa.2012.07.033 URL
[14]	VAN DER AALST W M P, VAN DONGEN B F, HERBST J, et al. Workflow mining: A survey of issues and approaches[J]. Data & Knowledge Engineering, 2003, 47(2): 237-267. doi: 10.1016/S0169-023X(03)00066-1 URL
[15]	VAN DER AALST W M P, WEIJTERS A J M M. Process mining: A research agenda[J]. Computers in Industry, 2004, 53(3): 231-244. doi: 10.1016/j.compind.2003.10.001 URL
[16]	VAN DER AALST W, WEIJTERS T, MARUSTER L. Workflow mining: Discovering process models from event logs[J]. IEEE Transactions on Knowledge and Data Engineering, 2004, 16(9): 1128-1142. doi: 10.1109/TKDE.2004.47 URL
[17]	陈好楠, 杨勇, 周同明, 等. 基于层次聚类的船舶建造分段运输日志分析[J]. 造船技术, 2020(4): 79-84.
	CHEN Haonan, YANG Yong, ZHOU Tongming, et al. Analysis of shipbuilding block assembly transportation log based on hierarchical clustering[J]. Marine Technology, 2020(4): 79-84.
[18]	RUSSELL S J, NORVIG P. Artificial intelligence: A modern approach[M]. 3rd ed. Upper Saddle River, New Jersey: Pearson Education, Inc., 2010.
[19]	BAUM L E, PETRIE T, SOULES G, et al. A maximization technique occurring in the statistical analysis of probabilistic functions of Markov chains[J]. The Annals of Mathematical Statistics, 1970, 41(1): 164-171. doi: 10.1214/aoms/1177697196 URL
[20]	VITERBI A. Error bounds for convolutional codes and an asymptotically optimum decoding algorithm[J]. IEEE Transactions on Information Theory, 1967, 13(2): 260-269. doi: 10.1109/TIT.1967.1054010 URL
[21]	RABINER L R. A tutorial on hidden Markov models and selected applications in speech recognition[J]. Proceedings of the IEEE, 1989, 77(2): 257-286. doi: 10.1109/5.18626 URL
[22]	BAUM L E, PETRIE T. Statistical inference for probabilistic functions of finite state Markov chains[J]. The Annals of Mathematical Statistics, 1966, 37(6): 1554-1563. doi: 10.1214/aoms/1177699147 URL

日期	分段	起点	终点	起点状态	终点状态
4月14日	101	G场	I场	组装	堆场周转
4月16日	101	I场	D场	堆场周转	预舾装
4月17日	101	D场	4#路	预舾装	堆场周转
4月17日	101	4#路	D场	堆场周转	预舾装
5月2日	101	D场	E场	预舾装	堆场周转
5月3日	101	E场	涂装4跨	堆场周转	冲砂打磨
5月4日	101	涂装4跨	涂装1跨	冲砂打磨	喷漆
5月11日	101	涂装1跨	K场	喷漆	堆场周转
6月17日	101	K场	C场	堆场周转	预总组
3月15日	102	组装1跨	A场	组装	堆场周转
…	…	…	…	…	…

序号	起点分段状态	终点分段状态的条件概率
序号	起点分段状态	堆场周转	组装	预舾装	冲砂打磨	喷漆	预总组
1	堆场周转	0.66	0.02	0.06	0.13	0.01	0.05
2	组装	0.54	0.43	0.01	0.01	0.00	0.00
3	预舾装	0.62	0.00	0.27	0.11	0.00	0.00
4	冲砂打磨	0.03	0.00	0.00	0.02	0.96	0.00
5	喷漆	0.89	0.00	0.00	0.01	0.03	0.03
6	预总组	0.07	0.00	0.00	0.00	0.00	0.30

序号	场地	分段状态的条件概率
序号	场地	堆场周转	组装	预舾装	冲砂打磨	喷漆	预总组
1	A场	0.13	0.17	0.24	0.00	0.00	0.00
9	G场	0.03	0.09	0.00	0.00	0.00	0.00
13	I场	0.09	0.04	0.00	0.00	0.00	0.00

分段状态	真实数/次	HMM¹				HMM²
分段状态	真实数/次	准确数/次	错判为	错判数/次	准确率/%	准确数/次	错判为	错判数/次	准确率/%
堆场周转	450	407	组装,预舾装, 预总组	17, 22, 4	90.4	419	预舾装,组装, 预总组	25, 2, 4	93.1
组装	121	120	堆场周转	1	99.2	118	堆场周转	3	97.5
预舾装	49	23	堆场周转, 组装	25, 1	46.9	31	堆场周转	18	63.3
冲砂打磨	59	59	—	—	100.0	59	—	—	100.0
喷漆	58	58	—	—	100.0	58	—	—	100.0
预总组	32	31	堆场周转	1	96.9	31	堆场周转	1	96.9
总计	769	698	—	71	90.8	716	—	53	93.1
分段状态	真实数/次	HMM³				HMM⁴
分段状态	真实数/次	准确数/次	错判为	错判数/次	准确率/%	准确数/次	错判为	错判数/次	准确率/%
堆场周转	450	418	预舾装,组装, 预总组	23, 5, 4	92.9	430	预舾装,组装, 预总组	15, 1, 4	95.6
组装	121	120	堆场周转	1	99.2	118	堆场周转	3	97.5
预舾装	49	24	堆场周转	25	49.0	24	堆场周转	25	49.0
冲砂打磨	59	57	喷漆	2	96.6	59	—	—	100.0
喷漆	58	58	—	—	100.0	58	—	—	100.0
预总组	32	31	堆场周转	1	96.9	30	堆场周转	2	93.8
总计	769	708	—	61	92.1	719	—	50	93.5

HMM	第一类非生产性转运/次	第二类非生产性转运/次	总转运/次
真实值	309	92	711
HMM¹	316	122	711
HMM²	311	107	711
HMM³	325	111	711
HMM⁴	326	89	711