广东工业大学学报 ›› 2022, Vol. 39 ›› Issue (05): 120-126,136.doi: 10.12052/gdutxb.220063

• • 上一篇    下一篇

基于深度神经网络的注塑过程预测控制

黎耀东, 任志刚, 吴宗泽   

  1. 广东工业大学 自动化学院，广东 广州 510006
  • 收稿日期:2022-03-29 发布日期:2022-07-18
  • 通信作者: 任志刚(1987−)，副教授，博士，硕士生导师, 主要研究方向为复杂工业过程优化与控制、工业智能，E-mail：renzhigang@gdut.edu.cn
  • 作者简介:黎耀东(1997−)，男，硕士研究生，主要研究方向为深度学习和模型预测控制，E-mail：2112004060@mail2.gdut.edu.cn
  • 基金资助:
    国家自然科学基金资助项目 (62073088, U1911401)；广东省重点领域研发计划项目(2021B0101200005)；广东省基础与应用基础研究基金资助项目(2019A1515011606)

Deep Neural Network Based Predictive Control for Injection Molding Process

Li Yao-dong, Ren Zhi-gang, Wu Zong-ze   

  1. School of Automation, Guangdong University of Technology, Guangzhou 510006, China
  • Received:2022-03-29 Published:2022-07-18

HTML

PDF (PC)

2140

摘要： 注塑机作为现代工业中重要的塑料件生产与制造设备，其智能化发展一直备受业界关注。伴随着航空航天、电力电子、汽车制造等行业的发展，如何实现对注塑件的高精度、高效率、绿色节能化生产，是目前注塑机的重要研制方向。本文针对注塑成型过程中采用传统模型预测控制(Model Predictive Control, MPC)难以保证跟踪控制实时性的问题，提出一种结合深度神经网络(Deep Neural Network, DNN)学习的注塑过程预测控制方法。在注塑机注射过程动力学模型基础上，创建带约束条件的模型预测控制器，对控制器运行数据进行采集并用以训练深度神经网络，实现了基于深度神经网络控制的注射速度的跟踪预测控制。仿真实验结果表明，采用本文所提出的学习预测控制策略能够有效避免注塑过程中因模型预测控制所产生的复杂计算，并满足工业实时性要求，具有应用前景。

关键词: 深度神经网络, 模型预测控制, 注塑成型, 最优控制

Abstract: The development of injection molding machines has always piqued the industry's interest as an important production and manufacturing equipment for plastic parts in modern industry. How to achieve high-precision, high-efficiency, green, and energy-saving injection molding parts is an important development direction for injection molding machines as the aerospace, power electronics, automobile manufacturing, and other industries grow. A strategy that combines the deep neural network (DNN) is proposed to realize the predictive control of the injection molding process of the injection molding machine, in response to the problem that traditional model predictive control (MPC) finds it difficult to guarantee real-time tracking in the injection molding process. A model predictive controller with constraints is created based on the dynamic model of the injection molding machine injection process, and the controller's operating data is collected and used to train the deep neural network to realize the predictive control of the injection molding process based on the deep neural network control. The simulation results show that this strategy can effectively avoid the complex calculation caused by the model predictive control in the injection molding process and meet the real-time requirements and has a broad application prospect.

Key words: deep neural network (DNN), model predictive control (MPC), injection molding process, optimal control

中图分类号: 

  • TP271
[1] WANG L M, LIU F F, YU J X, et al. Iterative learning fault-tolerant control for injection molding processes against actuator faults [J]. Journal of Process Control, 2017, 59(11): 59-72.
[2] ZOU T, WU S, ZHANG R D. Improved state space model predictive fault-tolerant control for injection molding batch processes with partial actuator faults using ga optimization [J]. ISA Transactions, 2018, 73(2): 147-153.
[3] 刘欣, 范希营, 郭永环, 等. 注塑工艺参数优化研究现状及发展趋势[J]. 塑料科技, 2021, 49(2): 106-110.
LIU X, FAN X Y, GUO Y H, et al. Research status and development trend of injection molding process parameter optimization [J]. Plastic Science and Technology, 2021, 49(2): 106-110.
[4] GAO F R, YANG Y, SHAO C. Robust iterative learning control with applications to injection molding process [J]. Chemical Engineering Science, 2001, 56(24): 7025-7034.
[5] TAN K K, TANG J C. Learning-enhanced PI control of ram velocity in injection molding machines [J]. Engineering Applications of Artificial Intelligence, 2002, 15(1): 65-72.
[6] 崔彩莲, 孙明轩. 注塑机注射速度的离散预期学习控制[J]. 机械工程学报, 2007, 43(6): 41-47.
CUI C L, SUN M X. Discrete anticipatory learning control of ram velocity in injection molding machines [J]. Chinese Journal of Mechanical Engineering, 2007, 43(6): 41-47.
[7] HOPMANN C, RESSMANN A, REITER M, et al. A self-optimising injection moulding process with model-based control system parameterisation [J]. International Journal of Computer Integrated Manufacturing, 2016, 29(11): 1190-1199.
[8] 陈胜裕, 李辉. 基于CNC人工蜂群算法的全电动注塑机料筒温控系统[J]. 塑料, 2021, 50(4): 65-70.
CHEN S Y, LI H. Temperature control system of full electric injection molding machine barrel based on CNC artificial bee colony algorithm [J]. Plastics, 2021, 50(4): 65-70.
[9] HOPMANN C, KAHVE C E, SCHMITZ M, et al. A comparison of predictive control strategies for a highly segmented injection mold tempering [J]. Polymer, 2021, 218(5): 123494.
[10] HONEK M, KVASNICA M, SZUCS A, et al. A low-complexity explicit MPC controller for AFR control [J]. Control Engineering Practice, 2015, 42(9): 118-127.
[11] SUARDI A, LONGO S, KERRIGAN E C, et al. Explicit MPC: hard constraint satisfaction under low precision arithmetic [J]. Control Engineering Practice, 2016, 47(5): 60-69.
[12] ZEILINGER M N, JONES C N, MORARI M. Real-time suboptimal model predictive control using a combination of explicit MPC and online optimization [J]. IEEE Transactions on Automatic Control, 2011, 56(7): 1524-1534.
[13] 郭心德, 丁宏强. 离散制造智能工厂场景的AGV路径规划方法[J]. 广东工业大学学报, 2021, 38(6): 70-76.
GUO X D, DING H Q. An AGV path planning method for discrete manufacturing smart factory [J]. Journal of Guangdong University of Technology, 2021, 38(6): 70-76.
[14] GUO Y M, LIU Y, OERLEMANS A, et al. Deep learning for visual understanding: a review [J]. Neurocomputing, 2016, 187(8): 27-48.
[15] Gary Yen, 栗波, 谢胜利. 地球流体动力学模型恢复的长短期记忆网络渐进优化方法[J]. 广东工业大学学报, 2021, 38(6): 1-8.
GARY Y, LI B, XIE S L. An evolutionary optimization of LSTM for model recovery of geophysical fluid dynamics [J]. Journal of Guangdong University of Technology, 2021, 38(6): 1-8.
[16] REN Z G, LAI J L, WU Z Z, et al. Deep neural networks-based real-time optimal navigation for an automatic guided vehicle with static and dynamic obstacles [J]. Neurocomputing, 2021, 443(9): 329-344.
[17] LOCKNER Y, HOPMANN C. Induced network-based transfer learning in injection molding for process modelling and optimization with artificial neural networks [J]. The International Journal of Advanced Manufacturing Technology, 2021, 112(11): 3501-3513.
[18] CHENG L, WANG Z, SONG Y, et al. Real-time optimal control for irregular asteroid landings using deep neural networks [J]. Acta Astronautica, 2020, 170(5): 66-79.
[19] LUCIA S, NAVARRO D, KARG B, et al. Deep learning-based model predictive control for resonant power converters [J]. IEEE Transactions on Industrial Informatics, 2020, 17(1): 409-420.
[20] CHEON K, KIM J, HAMADACHE M, et al. On replacing PID controller with deep learning controller for DC motor system [J]. Journal of Automation and Control Engineering, 2015, 3(6): 452-456.
[21] WANG K K, SHEN S F. Computer-aided design and fabrication of molds and computer control of injection molding [M]. New York: Cornell University, 1985: 1-50.
[22] KINGMA D P, BA J. Adam: a method for stochastic optimization[EB/OL]. (2017-01-30) [2022-04-16]. https://doi.org/10.48550/arXiv.1412.6980.
[23] LECUN Y, BENGIO Y, HINTON G. Deep learning [J]. Nature, 2015, 521(21): 436-444.
[24] KHAN A, SOHAIL A, ZAHOORA U, et al. A survey of the recent architectures of deep convolutional neural networks [J]. Artificial Intelligence Review, 2020, 53(8): 5455-5516.
[1] 马山, 汤超宇, 马骏峰, 彭涛. 基于对称交替方向乘子法的单列车最优运行控制[J]. 广东工业大学学报, 2022, 39(05): 38-45.
[2] 杨翊卓, 代伟. 事件触发机制下的工业过程多速率模型预测控制方法[J]. 广东工业大学学报, 2022, 39(05): 68-74.
[3] 彭积广, 肖涵臻. 模型预测控制下多移动机器人的跟踪与避障[J]. 广东工业大学学报, 2022, 39(05): 93-101.
[4] 王溥希, 冯广. 带有随机时延和丢包的网络控制系统最优控制策略研究[J]. 广东工业大学学报, 2014, 31(1): 32-35.
[5] 马静； 胡玉娥； 詹习生； . 基于模型预测的迭代学习控制器设计[J]. 广东工业大学学报, 2006, 23(2): 48-53.
[6] 王钦若； 张宁； 张振昭； . 一种多变量推理控制系统的设计方法[J]. 广东工业大学学报, 1999, 16(1): 13-16.
[7] 张平健； . 混合状态—控制约束下最优控制的必要条件[J]. 广东工业大学学报, 1997, 14(1): 14-20.
[8] 徐湘元. 最优控制的设计与研究[J]. 广东工业大学学报, 1996, 13(4): 71-76.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2017 《广东工业大学学报》编辑部
地址：广州市东风东路729号广东工业大学 邮编：510090
电话：020-37626653,020-37626139,020-37626396
邮箱：xbzrb@gdut.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发