简单形状零件拉深成形智能化控制技术研究

板材成形智能化是冲压工艺的一项新技术.以圆锥形件和锥壁盒形件为研究对象,分别讨论了实现实时监测、实时识别、实时预测和实时控制所需的关键技术,逐步形成了拉深成形智能化控制的理论基础.在实时识别和实时预测阶段采用人工神经网络,在压边力控制阶段采用基于破裂临界曲线的控制方法,能够得到较满意的结果.研究过程中开发了智能化拉深实验系统,实验证明识别、预测精度较好.

Intellectualization of controlling in deep drawing of simple shape workpieces

Taken the conical part and the rectangular box with conical sidewall as examples,the key technologies in their real-time monitoring,identifying,predicting and controlling were studied,which constitute the basis of the intelligent deep drawing control theory.The satisfying result was obtained when the artificial neural network was used in the case of real-time identification and prediction as well as when the load curve was along the critical rupture curve in the case of BHF(blank hold force) controlling.In this study,the testing system for intelligent deep drawing is trial-produced,in which the precision of identification and prediction is satisfying.