基于BP网络的注塑件翘曲变形工艺参数优化研究

利用BP神经网络对注塑工艺参数及其相对应的翘曲变形量样本进行训练，得到了描述工艺参数到翘曲量映射关系的人工神经网络（ANN）模型；验证了此模型的准确性；得出了工艺参数与注塑件翘曲变形量的内在联系，为以后的参数优化以及翘曲量预测起到重要的指导作用。

Optimization of Process Parameters of Warp Distortion of Plastics Parts Based on BP Network

Non-linear relationship between the injection process parameters and the degree of warp distortion was established through the neural network, which was testified by the test samples. The optimization of process parameters based on the neural network and orthogonal test was carried out. So, the internal links of process pammeters and the degree of warp distortion was found, which was an effective guidance to further optimize the process parameters and predict the degree of warp distortion.