应用神经网络进行薄板冲压件坯料优化  

Optimization of Initial Blank Shape of Sheet Metal by Using Neural Network

作　　者：黄玉萍[1] 阮锋[1] 

机构地区：[1]华南理工大学机械工程学院,广东广州510640

出　　处：《锻压装备与制造技术》2008年第1期57-59,共3页China Metalforming Equipment & Manufacturing Technology

基　　金：广东省工业攻关计划资助项目(2004A11403002)

摘　　要：坯料设计是冲压工艺设计的重要环节。传统的经验类比设计对于复杂拉深件难以奏效,单纯应用有限元增量模拟法进行"试错"耗时多、效率低。本文提出一种利用神经网络与有限元方法结合进行坯料优化的方法,通过实例验证了该方法的可行性,并且成功地将它应用于实际冲压件的毛坯设计。Initial blank shape design is one of the most important segments during processing design of sheet metal forming. Analogy method based on experience is insufficient for complex shape part and Trail-and-error.Only simple finite element increment simulation method is time consuming.One initial blank shape design method based on both neural network and FEM has been introduced in this text.The feasibility of this method has been proved by practice example ,which is successful used in reality design.

关 键 词：机械制造 薄板坯料 冲压 优化设计 神经网络 

分 类 号：TG386[金属学及工艺—金属压力加工]