基于神经网络和遗传算法的三维打印工艺参数优化

Optimization of the 3DP Printing Parameters Based on the Neural Networks and Genetic Algorithm

分析了粉末材料三维打印(three dimensional printing,3DP)过程中影响成型精度的因素,采用试验的方法确定打印过程中的三维制件的收缩率范围。以"H"型工件为标准,建立了基于神经网络(neural network,NN)的制件尺寸精度误差和打印工艺参数之间关系的模型。以制件最小尺寸误差为目标,采用遗传算法(genetic algorithm,GA)对3DP中的工艺参数如饱和度、层厚和X、Y、Z这3个方向的收缩补偿值进行优化,获得了相应的打印工艺参数。采用3DP默认的打印参数、打印参数的最小值、最大值以及NN-GA得到的参数进行对比试验。结果表明:采用NN-GA获得的工艺参数打印的制件的尺寸误差最小,可以预测3DP成型制件相对尺寸误差。

The factors affect the printing accuracy for powdery materials is analyzed in three dimensional printing（ 3DP） process,the experiment is used to determine the range of shrinkage of the printing process for the threedimensional components. The ＂ H＂ type component as the test specimen,the neural network（ NN） is used to describe the complicated relationship between the dimensional accuracy of component and printing processing parameters. As the goal of the minimum dimensional accuracy of specimen,the genetic algorithm（ GA） is used to optimize the 3DP printing parameters such as saturation,the layer thickness and compensation in three directions X,Y and Z respectively. The comparing experiments for 3DP using default parameters of printer,the minimum and maximum value in the range of printing parameters,and the NN-GA obtained parameters are conducted,and the results show that the dimensional accuracy is the best using the printing processing parameters of the NN-GA obtained,which show that the NN-GA can predict the dimensional accuracy for 3DP printing processing and provide the reference for other similar fabrication method.