摘要
【目的】为提高熔融沉积(fused deposition modeling,FDM)型3D打印机喷头温度控制中的精度和稳定性,提出了基于深度确定策略梯度(deep deterministic policy gradient,DDPG)算法的温度控制策略。【方法】首先,对3D打印机打印喷头进行数学建模;然后,在设计基于DDPG算法的马尔科夫决策过程(markov decision process,MDP)模型时,以目标温度、当前温度和时间作为状态观测变量,以比例积分微分(proportion integral differential,PID)控制器的3个参数作为输出动作值,设置多个全连接隐藏层来捕捉复杂非线性关系,并基于温度偏差设计了一组复合型奖励函数;最后,搭建仿真模型对算法进行了验证。【结果】与传统PID和模糊PID控制策略相比,基于DDPG的PID响应时间分别缩短了37.01%和18.36%,调节时间分别缩短了27.14%和8.91%,超调量分别降低了74.14%和58.89%。【结论】本方法显著提升了FDM型3D打印机喷头温度控制系统性能,可为3D打印喷头温度控制精度和稳定性研究提供参考。
[Objective]To improve the accuracy and stability in the temperature control of fused deposition modeling(FDM)3D print nozzle,a temperature control strategy was proposed on the basis of the deep deterministic policy gradient(DDPG)algorithm.[Method]First,a mathematical model of the 3D printer print-nozzle was established;then,when designing the markov decision process(MDP)model based on the DDPG algorithm,the target temperature,current temperature,and time were viewed as state observation variables,while three parameters of the proportion integral differential(PID)controller served as output action values,with several fully connected hidden layers deployed to capture complex nonlinear relationships and with a composite reward function designed on the basis of temperature deviation;finally,a simulation model was built for experimental verification.[Result]Compared with traditional PID and fuzzy PID control strategies,DDPG-based PID reduces the response time by 37.01%and 18.36%,the adjustment time by 27.14%and 8.91%,and the overshoot by 74.14%and 58.59%,respectively.[Conclusion]This method significantly enhances the performance of the temperature control system for FDM 3D print nozzle,providing a reference for studies on temperature control accuracy and stability of 3D print nozzle.
作者
谭平
周招
杨大胜
武永波
丁进
TAN Ping;ZHOU Zhao;YANG Dasheng;WU Yongbo;DING Jin(School of Automation and Electrical Engineering,Zhejiang University of Science and Technology,Hangzhou 310023,Zhejiang,China;School of Control Science and Engineering,Zhejiang University,Hangzhou 310007,Zhejiang,China)
出处
《浙江科技大学学报》
2025年第1期45-54,共10页
Journal of Zhejiang University of Science and Technology
基金
国家自然科学基金项目(52472424)
浙江省重点研发计划项目(2023C01243)。
作者简介
通信作者:武永波(1984—),男,河南省洛阳人,讲师,博士,主要从事物联网技术研究。E-mail:122145@zust.edu.cn。
