Please wait a minute...
Chin J Eng Design  2023, Vol. 30 Issue (1): 65-72    DOI: 10.3785/j.issn.1006-754X.2023.00.003
Design for Quality     
Study on fuzzy neural network PID stability control for extrusion force of ceramic slurry 3D printer
Jie YANG(),Zhuang-zhuang PENG,Shi-jie WANG,Cong MA,Long WANG,Guo-lin DUAN()
School of Mechanical Engineering, Hebei University of Technology, Tianjin 300401, China
Download: HTML     PDF(2976KB)
Export: BibTeX | EndNote (RIS)      

Abstract  

In view of the demand of extrusion force stability control in the process of micro-flow extrusion ceramic slurry 3D printer operation, according to the nonlinear and time-varying characteristics of the printer extrusion force control system, the advantages and disadvantages of the existing extrusion force stability control strategy were summarized, and the neural network structure was embedded in the fuzzy PID (proportional-integral-derivative) controller, so a fuzzy neural network PID stability control strategy for extrusion force was proposed. The strategy was based on a six-layer fuzzy neural network, with the extrusion force deviation value e and the deviation change rate ec as the input, and the PID controller control parameters as the output, to complete the forward fuzzy control process, and based on the self-learning advantage of the neural network to realize the reverse propagation and online update the neural network weight, to achieve the accurate adaptive adjustment of the extrusion force in the printing process. The Simulink simulation of extrusion force control, the extrusion force control experiment and the blank printing experiment showed that, compared with the traditional PID control strategy, the fuzzy neural network PID control strategy could reduce the overshoot by 20.9%, the extrusion force reached a stable state 90 s ahead of time, the pressure peak value decreased by 12 N, and the pressure valley value increased by 18 N; compared with the fuzzy PID control strategy, the overshoot was reduced by 1.73%, the extrusion force reached a stable state 56 s ahead of time, the pressure peak decreased by 4 N, and the pressure valley increased by 8 N; the fuzzy neural network PID control strategy had certain advantages, which could make the control precision of extrusion force higher, the stability speed faster, the overshoot smaller, the overall shape quality of the printed body better, and could also make the system more robust. The research results provide new ideas and methods for PID control and intelligent control of other industrial equipment.



Key words3D printing      extrusion force      stability control strategy      fuzzy neural network     
Received: 26 March 2022      Published: 06 March 2023
CLC:  TQ 174.6  
Corresponding Authors: Guo-lin DUAN     E-mail: yj18812616880@163.com;glduan@hebut.cn
Cite this article:

Jie YANG,Zhuang-zhuang PENG,Shi-jie WANG,Cong MA,Long WANG,Guo-lin DUAN. Study on fuzzy neural network PID stability control for extrusion force of ceramic slurry 3D printer. Chin J Eng Design, 2023, 30(1): 65-72.

URL:

https://www.zjujournals.com/gcsjxb/10.3785/j.issn.1006-754X.2023.00.003     OR     https://www.zjujournals.com/gcsjxb/Y2023/V30/I1/65


陶瓷浆料3D打印机挤压力模糊神经网络PID稳定控制研究

针对在微流挤出陶瓷浆料3D打印机作业过程中挤压力稳定控制的需求,根据打印机挤压力控制系统非线性、时变性的特点,总结了现有挤压力稳定控制策略的优缺点,并在模糊PID(proportion-integral-derivative,比例?积分?微分)控制器中嵌入神经网络结构,提出了挤压力模糊神经网络PID稳定控制策略。该策略基于六层模糊神经网络,以挤压力偏差值e和偏差值变化率ec为输入,PID控制器控制参数为输出,完成正向模糊控制过程,并基于神经网络的自学习优势实现反向传播及在线更新神经网络权值,以实现打印过程中挤压力的精准自适应调节。挤压力控制Simulink仿真、挤压力控制实验及坯体打印实验表明:相较于传统PID控制策略,采用模糊神经网络PID控制策略可使超调量减小20.9%,挤压力提前90 s达到稳定状态,压力峰值减小12 N,压力谷值增大18 N;相较于采用模糊PID控制策略,超调量减小1.73%,挤压力提前56 s达到稳定状态,压力峰值减小4 N,压力谷值增大8 N;模糊神经网络PID控制策略具有一定的优越性,可使打印过程中挤压力的控制精度更高,稳定速度更快,超调量更小,所打印坯体的整体形貌质量更优,也可使控制系统的鲁棒性更好。研究结果为其他工业设备的PID控制、智能控制提供了新的思路和方法。


关键词: 3D打印,  挤压力,  稳定控制策略,  模糊神经网络 
Fig.1 Extrusion force stability control principle
Fig.2 Membership functions of e and Δkp
eecΔkpΔkiΔkd
NVBNBPVBNVBNVB
NBNMPBNBNS
NMNVBPVBNVBPS
NSNMPMNMNS
ZOPSNSPSNS
?????
PBNBZOZOPB
PSPBNMPBZO
PMNSNSPSPS
NSPSZOZONS
NMPVBNMPSPS
Table 1 Fuzzy control rule table
Fig.3 Fuzzy neural network structure
Fig.4 Fuzzy neural network PID control model
Fig.5 Simulation results of extrusion force stability control under three control strategies
Fig.6 Extrusion force stability control process
Fig.7 Ceramic slurry 3D printer
Fig.8 Extrusion force control result
Fig.9 Morphology of ceramic body
[1]   刘文进,周国相,林坤鹏,等.基于浆料形态的陶瓷3D打印技术的浆料体系研究进展[J].硅酸盐通报,2021, 40(6):1918-1926.
LIU Wen-jin, ZHOU Guo-xiang, LIN Kun-peng, et al. Research progress on slurry system of ceramic 3D printing technology based on slurry morphology[J]. Bulletin of the Chinese Ceramic Society, 2021, 40(6): 1918-1926.
[2]   CAI Jia-wei, ZHANG Bai-cheng, ZHANG Mao-hang, et al. Indirect 3D printed ceramic: A literature review[J]. Journal of Central South University, 2021(4): 983-1002.
[3]   周婧,段国林,卢林,等.陶瓷浆料微流挤压成形关键问题研究[J].中国机械工程,2015,26(22):3097-3102. doi:10.3969/j.issn.1004-132X.2015.22.018
ZHOU Jing, DUAN Guo-lin, LU Lin, et al. Research on several key problems of microflow extrusion forming of ceramic slurry[J]. China Mechanical Engineering, 2015, 26(22): 3097-3102.
doi: 10.3969/j.issn.1004-132X.2015.22.018
[4]   NPKA B, DC B, MZA B. Optimization of 3D printing parameters of screw type extrusion (STE) for ceramics using the Taguchi method [J]. Ceramics International, 2019, 45(2): 2351-2360.
[5]   刘志鹏.基于微流挤出工艺的3D打印机控制系统研究[D]. 天津:河北工业大学, 2020:43-55. doi:10.13433/j.cnki.1003-8728.20190249
LIU Zhi-peng. Research on 3D printer control system based on microfluidic extrusion process[D]. Tianjin: Hebei University of Technology, 2020: 43-55.
doi: 10.13433/j.cnki.1003-8728.20190249
[6]   MASON M S, HUANG T, LANDERS R G, et al. Freeform extrusion of high solids loading ceramic slurries, part I: extrusion process modeling[C]//Seventeenth Annual Solid Freeform Fabrication Symposium, Rolla, Missouri, September 14-16, 2006.
[7]   焦盼德,李淑娟,杨磊鹏,等.功能梯度材料快速成形过程建模与控制[J].中国机械工程,2017,28(6):733-738. doi:10.3969/j.issn.1004-132X.2017.06.016
JIAO Pan-de, LI Shu-juan, YANG Lei-peng, et al. Modeling and control of rapid deposition for functionally gradient material components[J]. China Mechanical Engineering, 2017, 28(6): 733-738.
doi: 10.3969/j.issn.1004-132X.2017.06.016
[8]   李聪.基于STM32的陶瓷浆料3D打印机控制系统研究[D]. 天津:河北工业大学, 2021:47-55.
LI Cong. Research on the control system of ceramic slurry 3D printer based on STM32[D]. Tianjin: Hebei University of Technology, 2021: 47-55.
[9]   DIAZ-RODRIGUEZ I D, HAN S, BHATTACHARYYA S P, et al. Bookshelf: analytical design of PID controllers[J]. IEEE Control Systems Magazine, 2021, 41(1): 80-81.
[10]   DUTTA P, NAYAK S K. Grey wolf optimizer based PID controller for speed control of BLDC motor[J]. Journal of Electrical Engineering & Technology, 2021(2): 955-961.
[11]   AL-DHAIFALLAH M, KANAGARAJ N, NISAR K S. Fuzzy fractional-order PID controller for fractional model of pneumatic pressure system[J]. Mathematical Problems in Engineering, 2018(1): 1-9.
[12]   王伟,张晶涛,柴天佑.PID参数先进整定方法综述[J]. 自动化学报,2000(3):347-355.
WANG Wei, ZHANG Jing-tao, CHAI Tian-you. A survey of advanced PID parameters turning methods[J]. Acta Automatica Sinica, 2000(3): 347-355.
[13]   肖亚宁,孙雪,郭艳玲,等.基于BP-PID的选择性激光烧结温度控制系统设计[J].机床与液压,2021,49(24): 95-100. doi:10.3969/j.issn.1001-3881.2021.24.019
XIAO Ya-ning, SUN Xue, GUO Yan-ling, et al. Design of temperature control system based on BP-PID for selective laser sintering[J]. Machine Tool & Hydraulics, 2021, 49(24): 95-100.
doi: 10.3969/j.issn.1001-3881.2021.24.019
[14]   罗昌恩,张国林,戴毅.基于STM32小型四轴飞行器PID参数整定[J].电子世界,2018(19):37-38.
LUO Chang-en, ZHANG Guo-lin, DAI Yi. PID parameter tuning based on STM32 small quadcopter[J]. Electronics World, 2018(19): 37-38.
[15]   陈星.基于模糊神经网络PID控制的花茶烘焙温控系统设计[J].食品与机械,2020,36(9):131-137.
CHEN Xing. Design of tea baking temperature control system based on fuzzy neural network PID control[J]. Food & Machinery, 2020, 36(9): 131-137.
[16]   郭佳跃,韦根原.基于自适应神经网络模糊PID的磨煤机控制研究[J].热能动力工程,2022,37(2):148-154.
GUO Jia-yue, WEI Gen-yuan. Research on coal mill control based on adaptive neural network fuzzy PID[J]. Journal of Engineering for Thermal Energy and Power, 2022, 37(2): 148-154.
[17]   DAVANIPOUR M, JAVANMARDI H, GOODARZI N. Chaotic self-tuning PID controller based on fuzzy wavelet neural network model[J]. Iranian Journal of Science and Technology, Transactions of Electrical Engineering, 2018(3): 357-366.
[18]   PARK D, LE T L, QUYNH N V, et al. Online tuning of PID controller using a multilayer fuzzy neural network design for quadcopter attitude tracking control[J]. Frontiers in Neurorobotics, 2021,14: 619350.
[19]   杨艺,虎恩典.基于S函数的BP神经网络PID控制器及Simulink仿真[J].电子设计工程,2014,22(4):29-31. doi:10.3969/j.issn.1674-6236.2014.04.009
YANG Yi, HU En-dian. Simulink simulation of BP neural network PID controller based on S-function[J]. Electronic Design Engineering,2014,22(4):29-31.
doi: 10.3969/j.issn.1674-6236.2014.04.009
[20]   李绍铭,赵伟.基于S函数的RBF神经网络PID控制器Simulink仿真[J]. 安徽冶金科技职业学院学报,2008(1): 19-21. doi:10.3969/j.issn.1672-9994.2008.01.007
LI Shao-ming, ZHAO Wei. Simulink simulation of RBF network PID controlle based on S-function [J]. Journal of Anhui Vocational College of Metallurgy and Technology, 2008(1): 19-21.
doi: 10.3969/j.issn.1672-9994.2008.01.007
[1] Bowei XIE,Mohui JIN,Zhou YANG,Jieli DUAN,Mingyu QU,Jinhui LI. Research on mechanical properties and model parameters of 3D printed TPU material[J]. Chin J Eng Design, 2023, 30(4): 419-428.
[2] Jing ZHOU,Hong-fei GAO,Lin LU,Guo-lin DUAN. Research on extrusion force of micro-flow extrusion 3D printer controlled by variable universe fuzzy PID[J]. Chin J Eng Design, 2022, 29(5): 572-578.
[3] CAO Wen-yi, CHEN Ji-min, YUAN Yan-ping, LIU Si-da. Development of FDM 3D printer based on hybrid coordinate system[J]. Chin J Eng Design, 2019, 26(1): 110-115.
[4] CAO Yan-Long, LIU Bao-Jun, YANG Jiang-Xin. New cost-tolerance function based on fuzzy neural networks[J]. Chin J Eng Design, 2004, 11(2): 73-76.