扰动累积下基于机器学习的重调度方式选择

中国机械工程

扰动累积下基于机器学习的重调度方式选择

唐秋华1,2,3;成丽新1,2,3;张利平1,2,3

1.武汉科技大学生产系统工程研究所，武汉，430081
2.武汉科技大学冶金装备及其控制教育部重点实验室，武汉，430081
3.武汉科技大学机械传动与制造工程湖北省重点实验室，武汉，430081

出版日期:2019-02-25 发布日期:2019-02-26
基金资助:
国家自然科学基金资助项目(51875421，51875420)

Rescheduling Mode Selection under Recessive Disturbance Accumulation via Machine Learning

TANG Qiuhua1,2,3;CHENG Lixin1,2,3 ;ZHANG Liping1,2,3

1.Institute of Production Systems Engineering, Wuhan University of Science and Technology, Wuhan, 430081
2.Key Laboratory of Metallurgical Equipment and Control Technology, Wuhan University of Science and Technology, Wuhan, 430081
3.Hubei Key Laboratory of Mechanical Transmission and Manufacturing Engineering, Wuhan University of Science and Technology, Wuhan, 430081

Online:2019-02-25 Published:2019-02-26

摘要/Abstract

摘要： 针对扰动累积下重调度方案甄选问题，提出了一种基于数据学习的重调度方式选择方法。利用累积误差时间来量化隐性扰动，用数据反映实时生产加工状况；构建扰动累积下重调度模型，基于遗传仿真获取评价重调度决策的因素、决策标签。分析仿真样本特征，并基于数据特征降维映射，构建基于支持向量机的分类决策模型以学习生产状况与重调度方式的内在联系，从而帮助生产管理者快速制定决策，提高响应速度。实验验证了所建立重调度方式选择框架的合理性和有效性。

关键词: 重调度方式选择, 遗传仿真, 降维, 支持向量机分类

Abstract: A method of rescheduling mode selection was proposed to solve the problems of disturbance cumulative rescheduling scheme selection based on data learning. The cumulative disturbance time was used to quantify the implicit disturbance, and the real-time production and processing status were reflected by data. The cumulative disturbance rescheduling model was constructed, and the factors and decision labels of rescheduling decision were obtained based on genetic simulation. The characteristics of the simulation samples were analyzed, and based on the data feature reduction mapping, the SVM-based classification decision model was constructed to learn the internal relationship between the production situation and the rescheduling mode, which would help the production managers to make the decision quickly and improve the response speed. Finally, the rationality and effectiveness of the rescheduling selection framework were verified by experiments.

Key words: rescheduling mode selection, genetic simulation, dimension reduction, support vector machine(SVM) classification

中图分类号:

TP181

唐秋华1,2,3;成丽新1,2,3;张利平1,2,3. 扰动累积下基于机器学习的重调度方式选择[J]. 中国机械工程.

TANG Qiuhua1,2,3;CHENG Lixin1,2,3 ;ZHANG Liping1,2,3. Rescheduling Mode Selection under Recessive Disturbance Accumulation via Machine Learning[J]. China Mechanical Engineering.

参考文献

[1]KATRAGJINIK, VALLADA E, RUIZ R. Flow Shop Rescheduling under Different Types of Disruption[J]. International Journal of Production Research, 2013,51(3):780-797.
[2]WANG C, JIANG P Y. Manifold Learning Based Rescheduling Decision Mechanism for Recessive Disturbances in RFID-driven Job Shops[J]. Journal of Intelligent Manufacturing, 2016:1-16.
[3]PETROVIC D, DUENAS A. A Fuzzy Logic Based Production Scheduling/Rescheduling in the Presence of Uncertain Disruptions[J]. Fuzzy Sets and Systems, 2006, 157(16): 2273-2285.
[4]乔非,李莉,王遵彤,等.面向半导体生产的改进的混合重调度策略研究[J].计算机集成制造系统,2007,13 (3): 558-562.
QIAO Fei, LI Li, WANG Zuntong, et al. Improved Hybrid Rescheduling Strategy for Semiconductor Manufacturing[J]. Computer Integrated Manufacturing System, 2007,13(3): 558-562.
[5]单晖.基于实时工况的Job Shop动态随机重调度方法研究[D].合肥：合肥工业大学,2008.
SHAN Hui. Research on Job Shop Dynamic Random Rescheduling Method Based on Real-time Operating Conditions [D]. Hefei : Hefei University of Technology,2008.
[6]陈静云.车间物联网数据采集关键技术研究[D]. 南京: 南京航空航天大学,2014.
CHEN Jingyun. Research on Key Technologies of Data Acquisition of Internet of Things in Workshop[D]. Nanjing : Nanjing University of Aeronautics and Astronautics, 2014.
[7]JIANG P, CAO W. An RFID-driven Graphical Formalized Deduction for Describing the Time-sensitive State and Position Changes of Work-in-progress Material Flows in a Job-shop Floor[J]. Journal of Manufacturing Science and Engineering, 2013,135(3): 189-197.
[8]刘明周,单晖,蒋增强,等.不确定条件下车间动态重调度优化方法[J].机械工程学报,2009,45(10):137-142.
LIU Mingzhou, SHAN Hui, JIANG Zengqiang, et al. Dynamic Rescheduling Optimization of Job-shop under Uncertain Conditions[J]. Journal of Mechanical Engineering, 2009,45(10): 137-142.
[9]刘壮,张中敏,杜先军. 基于改进 TOPSIS 的制造车间重调度决策方法研究[J].组合机床与自动化加工技术, 2017 (1):157-160.
LIU Zhuang, ZHANG Zhongmin, DU Xianjun. Rescheduling Decision Method of Manufacturing Shop Based on Improved TOPSIS[J]. Modular Machine Tool and Automatic Machining Technology, 2017 (1): 157-160.
[10]AKKAN C. Improving Schedule Stability in Single-machine Rescheduling for New Operation Insertion[J]. Computers & Operations Research, 2015, 64: 198-209.
[11]毕达天,邱长波,张晗.数据降维技术研究现状及其进展[J].情报理论与实践,2013,36(2):125-128.
BI Datian, QIU Changbo, ZHANG Han. Research Status and Progress of Data Dimension Reduction Technology[J]. Information Theory and Practice,2013,36(2):125-128.
[12]刘萌萌,刘亚岚,孙国庆,等.结合纹理特征的SVM样本分层土地覆盖分类[J].遥感技术与应用,2014,29(2):315-323.
LIU Mengmeng, LIU Yalan, SUN Guoqing,et al. Stratified Land Cover Classification Based on SVM Samples with Texture Characteristics[J]. Remote Sensing Technology and Application,2014,29(2):315-323.