摘要
为提高3D打印成型工件精度,降低其对不可控因子变化的敏感性,提出一种基于高斯过程回归建模和改进粒子群寻优的稳健参数设计方法。首先,选择4个显著的可控因子:层厚、打印速度、热床温度、喷嘴温度,并确定其可行域;其次,利用超拉丁方设计获取建模样本集,采用高斯过程回归分别拟合打印工件精度均值和方差的两个响应曲面模型;然后,以均值方差之和最小化为寻优目标,采用惯性权重非线性递减的改进粒子群算法对模型寻优;最后,通过现场实验对寻优结果进行验证。与传统双响应曲面法的对比表明,所提方法在优化结果、拟合性能、寻优能力等方面有显著的改善和提高。
In order to improve the precision of 3D printing molding workpiece and reduce the sensitivity to the change of uncontrollable factors,a robust parameter design method based on Gaussian process regression(GPR)modeling and improved particle swarm optimization was proposed.Firstly,four significant controllable factors were determined:layer thickness,printing speed,hot bed temperature and nozzle temperature,and the feasible region was determined.Secondly,the Latin hypercube sampling(LHS)was used to obtain the modeling sample set,and Gaussian process regression was used to fit the two response surface models of the accuracy mean and variance of the printed workpiece respectively.Then,minimizing the sum of mean and variance as the optimization goal,the improved particle swarm optimization was used to optimize the model parameters.Finally,the optimization result was verified through field experiments.Compared with the traditional dual response surface methodology,the proposed method can significantly improve the optimization result,fitting performance and searching ability.
作者
崔庆安
朱傲泉
CUI Qingan;ZHU Aoquan(School of Management Engineering,Zhengzhou University,Zhengzhou,Henan 450001,China;School of Economics&Management,Shanghai Maritime University,Shanghai 201306,China)
出处
《工业工程与管理》
北大核心
2022年第5期127-135,共9页
Industrial Engineering and Management
基金
国家自然科学基金资助项目(71571168,U1904211)。
关键词
稳健参数设计
3D打印
双响应曲面
高斯过程回归
改进粒子群算法
robust parameter design
3D printing
dual response surface methodology
Gaussian process regression
improved particle swarm optimization
