探讨了冷精密塑性加工模具设计过程中回弹量的影响因素 ,提出了一种基于神经网络技术的冷精密塑性加工回弹量模糊预测模型。应用具有更高智能的神经网络技术实现了对抽象模糊规则的自动纠错及记忆 ,符合人类学习模式 ,能合理表达经验知识 ,而且更能明确地表达模型的输入输出关系。 The influencing factors of springback in the process of cold precision plastic mold design were discussed. A fuzzy prediction model of springback in cold precision plastic processing based on neural network technology was proposed. The application of neural network technology with higher intelligence has realized the automatic error correction and memory of abstract fuzzy rules, which is in line with human learning mode, can reasonably express empirical knowledge, and can more clearly express the input-output relationship of the model.