陶瓷浆料3D打印机挤压力模糊神经网络PID稳定控制研究  

Study on fuzzy neural network PID stability control for extrusion force of ceramic slurry 3D printer

作　　者：杨杰[1] 彭壮壮 王世杰 马聪 王龙 段国林[1] YANG Jie;PENG Zhuang-zhuang;WANG Shi-jie;MA Cong;WANG Long;DUAN Guo-lin(School of Mechanical Engineering,Hebei University of Technology,Tianjin 300401,China)

机构地区：[1]河北工业大学机械工程学院,天津300401

出　　处：《工程设计学报》2023年第1期65-72,共8页Chinese Journal of Engineering Design

基　　金：中央引导地方科技发展资金资助项目(216Z1804G)。

摘　　要：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。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 intel

关 键 词：3D打印 挤压力 稳定控制策略 模糊神经网络 

分 类 号：TQ174.6[化学工程—陶瓷工业]