离散型生产车间以多品种、小批量的生产模式为主要特征，其按计划调度管控方案进行产品生产时，存在着大量的信息交互、生产异常、资源协调等问题。生产车间异常事件的出现更是增加了以计划调度为关键的车间管控难度，因此如何实现离散型生产车间异常事件预测与管控是企业在进行生产时亟需解决的难题。本文针对离散型生产过程的实际需求，从建立生产车间异常事件预测模型和设计有效的计划调度管控方法两方面出发，系统地研究生产车间异常事件预测与管控方法，为实现生产过程智能管控提供一种有效途径。具体研究内容如下：
（1）构建离散型生产车间异常事件预测与管控机制。在对生产车间异常事件分类的基础上，分析传统异常事件的管控流程及存在问题。将数字孪生技术、预测方法和计划调度管控方法进行融合，从生产前和生产中两个阶段出发，综合构建基于数字孪生的生产车间异常事件预测与管控运行机制。阐述运行机制的主要组成模块，并建立详细的异常事件预测与计划调度管控流程，从而解决生产车间管控机制不全面、动态响应能力不足、时效性差等问题。
（2）研究基于灰色马尔科夫的生产车间异常事件预测方法。建立生产车间异常事件灰色预测模型，对灰色预测结果的相对误差进行状态划分，并建立相应的马尔科夫预测模型对灰色预测结果进行修正，从而实现基于灰色马尔科夫的生产车间异常事件预测，为计划调度系统提供实时信息。
（3）建立生产车间异常事件预测下的车间计划调度管控模型。对车间计划调度管控问题进行分析，提出以最小化最大完工时间和总误期时间为目标的计划调度管控模型。为求解多目标计划调度管控模型，对NSGA-II算法进行改进，设计基于改进NSGA-II的计划调度管控算法主要结构。在此基础上，提出基于AHP的车间计划调度管控方案决策方法。
（4）以A企业为实例对象，首先运用生产车间异常事件预测与管控机制中的思想，进行车间计划调度管控分析和优化，运用算法求解得到初始计划调度管控方案。接着以车间设备故障数据为基础，基于灰色马尔科夫预测模型实现设备故障间隔时间预测。最后将设备故障预测数据反馈至计划调度系统，得到设备故障预测下的车间计划调度管控方案，据此可对车间进行重新调度，实现对生产车间异常事件的预测与管控。

The discrete production workshop is characterized by a multi-variety, small-lot production model. When the production is carried out according to the planning and scheduling control scheme, there are many problems such as information interaction, production exceptions, and resource coordination. The emergence of abnormal events in the production workshop increases the difficulty of workshop control with planning and scheduling as the key, so how to achieve the prediction and control of abnormal events on the discrete production floor is a challenge that needs to be solved by enterprises in production. In this paper, the actual needs of the discrete production process are addressed. A systematic study of the production workshop abnormal event prediction and control method is undertaken by establishing a production plant abnormal events prediction model and designing a practical planning and scheduling control method to provide an effective way to achieve intelligent control of the production process. The specific research content is as follows.

(1) Building an abnormal events prediction and control mechanism for discrete production workshops. Based on the classification of abnormal events on the production floor, the control process and problems of traditional abnormal events are analysed. By integrating digital twin technology, forecasting methods and planning and scheduling control methods, the digital twin-based production workshop abnormal events forecasting and control mechanism is built from two stages: pre-production and in-production. The main components of the operational mechanism are described, and a detailed abnormal events prediction and planning and scheduling control process is established to solve the problems of incomplete control mechanism, insufficient dynamic response capability and poor timeliness of the production workshop.

(2) Research on Grey-Markov based production workshop abnormal events prediction method. A grey prediction model for abnormal events on the production floor is established, the grey prediction results are divided into states, and the corresponding Markov prediction model is established to correct the grey prediction results, so as to realise Grey-Markov based abnormal events prediction on the production floor and provide real-time information for the planning and scheduling system.

(3) Develop a workshop planning and scheduling control model with abnormal events prediction on the production floor. The shop floor planning and scheduling control problem is analysed. A planning and scheduling control model with the objective of minimising the maximum completion time and the total delay time is proposed. In order to solve the multi-objective scheduling control model, the NSGA-II algorithm is improved, and the main structure of the improved NSGA-II-based scheduling control algorithm is designed. Based on this, the AHP-based workshop planning and scheduling control scheme decision method is proposed.

(4) Using enterprise A as an example object, the analysis and optimisation of workshop planning and scheduling control are first carried out using ideas from the production workshop abnormal event prediction and control mechanism. The initial planning and scheduling control solution is obtained using algorithmic solutions. Then, based on the workshop equipment failure data, a Grey-Markov prediction model is used to predict the interval between equipment failures. Finally, the equipment failure prediction data is fed back to the planning and scheduling system to obtain the workshop planning and scheduling control scheme under the equipment failure prediction. The workshop can be rescheduled to achieve the prediction and control of abnormal events in the production workshop.

[1]饶运清,刘世平,李淑霞,等.敏捷化车间制造执行系统研究[J].中国机械工程, 2002(08): 30-32+4．

[2]尹超,高玲,李孝斌.基于复杂网络理论的离散制造车间生产物流网络建模与分析[J].计算机集成制造系统, 2020, 26(08): 2157-2169.

[3]黄绍广.企业生产运行异常管理信息化系统研究[J].中国标准化, 2019(22): 220-221.

[4]Coalition S M L. Implementing 21st century smart manufacturing: workshop summary report. 2011.

[5]彭席席.欧美制造业强势回归对“中国制造2025”的若干启示[J].现代营销, 2016, 6: 18.

[6]周晓晓.我国智能制造的发展战略和体制保障[D].天津:天津师范大学, 2018.

[7]“十四五”智能制造发展规划专家系列解读[J].智能制造, 2022(01): 9.

[8]李炯.基于精益理念的生产过程控制优化[D].武汉:湖北工业大学, 2016.

[9]王进富,邵景峰,马晓红,等.织造过程异常事件的预警方法与系统设计[J].纺织学报, 2013, 34(05): 121-126.

[10]陶飞,程颖,程江峰,等.数字孪生车间信息物理融合理论与技术[J].计算机集成制造系统, 2017, 23(8): 1603-1611.

[11]Chakraborty S, Tah D. Real time statistical process advisor for effective quality control[J]. Decision Support Systems, 2006, 42(2): 700-711.

[12]马力,王奕娇,夏吉庆.企业管理中生产异常处理及改善方法探究[J].东北农业大学学报(社会科学版), 2009, 7(01): 82-84.

[13]Shi X J, Li F T, Zhang Y, et al. Quality control and tracing system of automotive airbag flexible assembly line[C]//Key Engineering Materials. Trans Tech Publications Ltd, 2011, 455: 66-71.

[14]林伏涛,谢元展,詹圣泽,等.制造业生产异常的诊断及其管理措施的构建[J].改革与战略, 2014, 30(10): 103-111.

[15]詹圣泽.制造企业生产异常诊断及控制研究[J].企业研究, 2015(02): 58-65.

[16]刘韬.企业管理中生产异常的处理及改善方法研究[J].企业改革与管理, 2019(14): 28.

[17]Breiter S, Arlinghaus J C. Disruption attributes for low-volume, complex product assembly[J]. Procedia CIRP, 2021, 104: 1710-1715.

[18]黄绍广.企业生产运行异常管理信息化系统研究[J].中国标准化, 2019(22): 220-221.

[19]王秋明,刘科成,高慧颖.基于灰色理论与BP神经网络的工序质量预测研究[J].北京理工大学学报, 2011, 31(02): 249-252.

[20]任莹晖,黄向明,马忠凯,等.基于信息熵的物料配送时间节点预测方法[J].中国机械工程, 2018, 29(22): 2725-2732.

[21]Wang W J, Sun H L, Lin S Z, et al. Failure Prediction and Intelligent Troubleshooting in the Application of the Coin Cell Production Line[J]. International Journal of Advancements in Computing Technology, 2012, 4(6).

[22]李亚杰,何卫平,董蓉,等.基于制造执行系统数据采集的工时预测与进化[J].计算机集成制造系统, 2013, 19(11): 2810-2818.

[23]尹超,甘德文,梁忠权,等.复杂机电产品关键装配工序物料质量损失评估及预警方法[J].计算机集成制造系统, 2014, 20(06): 1432-1442.

[24]闪四清,毛中慧,辛腾龙.基于BP神经网络的飞机装配车间异常事件预警[J].航空制造技术, 2014, 452(8): 42-47.

[25]杨志杰,郭顺生,郭钧.基于灰色多元线性回归的纺机行业订单预测[J].武汉理工大学学报(信息与管理工程版),2015,37(02):255-258.

[26]马超,徐迭石,张淑丽,等.大数据环境下离散制造车间异常事件发现方法[J].计算机应用与软件, 2017, 34(9): 288-293.

[27]Brahim I B, Addouche S A, El Mhamedi A, et al. Build a Bayesian Network from FMECA in the Production of Automotive Parts: Diagnosis and Prediction[J]. IFAC- PapersOnLine, 2019, 52(13): 2572-2577.

[28]姜天.云制造环境下车间生产异常预测与溯源方法研究[D].哈尔滨:哈尔滨理工大学, 2021.

[29]Bi K P, Tang N, Yan G H, et al. Research on Certain Type Equipment Failure Intervals Forecasting Based on Gray-Markov Model[C]//2019 International Conference on Quality, Reliability, Risk, Maintenance, and Safety Engineering (QR2MSE). IEEE, 2019: 21-27.

[30]Wang W P, Chen Z. A neuro-fuzzy based forecasting approach for rush order control applications[J]. Expert Systems with Applications, 2008, 35(1-2): 223-234.

[31]李俊青,潘全科,王玉亭.多目标柔性车间调度的Pareto混合禁忌搜索算法[J].计算机集成制造系统, 2010, 16(07): 1419-1426.

[32]宋存利.求解多目标混合流水车间调度的改进NSGAII算法[J/OL].计算机集成制造系统: 1-19[2020-11-23].

[33]Yuan M, Li Y, Zhang L, et al. Research on intelligent workshop resource scheduling method based on improved NSGA-II algorithm[J]. Robotics and Computer-Integrated Manufacturing, 2021, 71: 102141.

[34]纪明,赵骥,姜开宇.面向单件小批量生产的车间分层调度算法[J].模具工业, 2019, 45(03): 1-5.

[35]董海,徐晓鹏,谢谢.多目标优化算法求解多柔性作业车间调度问题[J].计算机科学, 2020, 47(12): 239-244.

[36]Goyal B, Kaur S. Minimizing waiting time of jobs in flow-shop scheduling: A heuristic approach[J]. Materials Today: Proceedings, 2020.

[37]张维存,李悦悦,左天帅.考虑人员装卸时间的作业车间调度问题[J].制造业自动化, 2020, 42(08): 87-94.

[38]宫敬程,庄存波,刘检华,等.基于磷虾群-禁忌算法的离散型装配调度问题[J].现代制造工程, 2020(09): 32-39+61.

[39]罗艳媚.基于改进蚁群算法求解双目标流水车间调度问题[J].桂林航天工业学院学报, 2020, 25(03): 353-358.

[40]Sun Y, Gong Q, Hu M, et al. Multi-Objective Optimization of Workshop Scheduling with Multiprocess Route Considering Logistics Intensity[J]. Processes, 2020, 8(7): 838.

[41]Bewoor L A, Prakash V C, Sapkal S U. Production scheduling optimization in foundry using hybrid Particle Swarm Optimization algorithm[J]. Procedia Manufacturing, 2018, 22: 57-64.

[42]Grieves M. Product lifecycle management: driving the next generation of lean thinking. McGraw-Hill, New York, 2006.

[43]Uhlemann T H J, Schock C, Lehmann C, et al. The digital twin: Demonstrating the potential of real time data acquisition in production systems[J]. Procedia Manufacturing, 2017, 9: 113-120.

[44]陶飞,张萌,程江峰,等.数字孪生车间——一种未来车间运行新模式[J].计算机集成制造系统, 2017, 23(01): 1-9.

[45]庄存波,刘检华,熊辉,等.产品数字孪生体的内涵、体系结构及其发展趋势[J].计算机集成制造系统, 2017, 23(04): 753-768.

[46]陶飞,刘蔚然,张萌,等.数字孪生五维模型及十大领域应用[J].计算机集成制造系统, 2019, 25(01): 1-18.

[47]Zhuang C, Liu J, Xiong H. Digital twin-based smart production management and control framework for the complex product assembly shop-floor[J]. The international journal of advanced manufacturing technology, 2018, 96(1): 1149-1163.

[48]Qi Q, Tao F, Hu T, et al. Enabling technologies and tools for digital twin[J]. Journal of Manufacturing Systems, 2021, 58: 3-21.

[49]刘志峰,陈伟,杨聪彬,等.基于数字孪生的零件智能制造车间调度云平台[J].计算机集成制造系统, 2019, 25(06): 1444-1453.

[50]Qiao Q, Wang J, Ye L, et al. Digital twin for machining tool condition prediction[J]. Procedia CIRP, 2019, 81: 1388-1393.

[51]武颖,姚丽亚,熊辉,等.基于数字孪生技术的离散型装配过程质量管控方法[J].计算机集成制造系统, 2019, 25(06): 1568-1575.

[52]宋悦,时祎瑜,于劲松,等.基于数字孪生的光电探测系统性能预测[J].计算机集成制造系统, 2019, 25(06): 1559-1567.

[53]Luo W, Hu T, Ye Y, et al. A hybrid predictive maintenance approach for CNC machine tool driven by Digital Twin[J]. Robotics and Computer-Integrated Manufacturing, 2020, 65: 101974.

[54]Zhang M, Tao F, Nee A Y C. Digital twin enhanced dynamic job-shop scheduling[J]. Journal of Manufacturing Systems, 2021, 58: 146-156.

[55]Negri E, Berardi S, Fumagalli L, et al. MES-integrated digital twin frameworks[J]. Journal of Manufacturing Systems, 2020, 56: 58-71.

[56]Wang Y, Wu Z. Model construction of planning and scheduling system based on digital twin[J]. The International Journal of Advanced Manufacturing Technology, 2020, 109(7): 2189-2203.

[57]赵阳,伏晓露,廖庆妙,等.基于数字孪生的智能脉动管控[J].航空制造技术, 2020, 63(Z1): 14-20.

[58]郭具涛,洪海波,钟珂珂,等.基于数字孪生的航天制造车间生产管控方法[J].中国机械工程, 2020, 31(07): 808-814.

[59]Liu Z, Chen W, Zhang C, et al. Intelligent scheduling of a feature-process-machine tool supernetwork based on digital twin workshop[J]. Journal of manufacturing systems, 2021, 58: 157-167.

[60]王昊琪,李浩,文笑雨,等.基于数字孪生的产品设计过程和工作量预测方法[J].计算机集成制造系统, 2022, 28(01): 17-30.

[61]张连超,刘蔚然,程江峰,等.卫星总装数字孪生车间物料准时配送方法[J].计算机集成制造系统, 2020, 26(11): 2897-2914.

[62]张映锋,赵曦滨,孙树栋.面向物联制造的主动感知与动态调度方法[M].北京:科学出版社, 2015.

[63]邓聚龙.灰色预测与决策[M].武汉:华中理工大学出版社, 1986.

[64]Zhan-Li M, Jin-Hua S. Application of Grey-Markov model in forecasting fire accidents[J]. Procedia Engineering, 2011, 11: 314-318.

[65]Wang Y, Chen C. Grey markov model forecast in economic system under incomplete information and its application on foreign direct investment[C]//2011 International Conference on Information Management, Innovation Management and Industrial Engineering. IEEE, 2011, 2: 117-120.

[66]李名升,任晓霞,佟连军.马尔科夫链在环境质量预测中的应用[J].环境工程, 2007(06): 78-80+100+5.

[67]Deb K, Pratap A, Agarwal S, et al. A fast and elitist multiobjective genetic algorithm: NSGA-II[J]. IEEE transactions on evolutionary computation, 2002, 6(2): 182-197.

[68]曹海涛.基于改进NSGA-Ⅱ的多目标柔性作业车间调度问题研究[D].杭州:浙江工业大学, 2019.

[69]Huo D X, Xiao X J, Pan Y J. MULTI-OBJECTIVE ENERGY-SAVING JOB-SHOP SCHEDULING BASED ON IMPROVED NSGA-II[J]. International Journal of Simulation Modelling (IJSIMM), 2020, 19(3).