论文部分内容阅读
针对激光烧结增材制造过程中出现的几何尺寸误差问题,采用正交实验与测量的方法获得训练样本,依据广义回归神经网络,建立了选择性激光烧结过程中工艺参数与成形收缩率之间的定量模型,以预测收缩率。定性分析了预热温度与支撑厚度对收缩率的影响,得到了各因素对收缩率影响的权重,并分析了主要因素间的交互作用。通过定性分析与定量预测,可为烧结过程中优化控制收缩提供一个新思路。
In order to solve the problem of geometrical dimension error during laser sintering additive manufacturing process, orthogonal test and measurement method were used to obtain training samples. According to generalized regression neural network, the relationship between process parameters and forming shrinkage rate during selective laser sintering Quantitative models to predict shrinkage. The influence of preheating temperature and supporting thickness on shrinkage was qualitatively analyzed. The weight of each factor on shrinkage was obtained, and the interaction among the main factors was analyzed. Through qualitative analysis and quantitative prediction, it can provide a new idea for optimal control of shrinkage during sintering.