作　　者： ; ; ; ; ; ;

机构地区： 燕山大学机械工程学院

出　　处： 《塑性工程学报》 2008年第6期58-61,共4页

摘　　要： 在板材成形智能化控制过程中,材料参数实时识别和最优工艺参数的实时预测是最重要的两个要素,实时识别时间的长短与实时预测精度的高低,直接影响到板材成形智能化控制的成败。文章利用神经网络技术实现了帽形件弯曲成形智能化控制过程中材料参数识别与最优工艺参数的预测,其材料参数及工艺参数均在数值模拟和试验提供的数据范围内,识别和预测模型的收敛精度均达到了0.1%,识别和泛化精度较高,满足帽形件弯曲智能化控制的要求。 The real-time identification of the material performance parameter and the real-time prediction of the optimum technical parameter are the two most important factors during the intellectualized control of sheet forming, the time of real-time identification and the precision of real-time prediction straightly influence the success of the intellectualized control of sheet forming. In this paper, the neural network technology was used to realize the identification of the material parameter and the prediction of the optimum technical parameter during the process of cap bending intellectualized control, the input data was provided by the two methods of numerical analysis and experiment, the identification model and the prediction model restraining precision has achieved 0. 1% in the data scope provided by numerical analysis and experiment, the precision is enough high to satisfy the request of cap bending intellectualized control.

关 键 词： 帽形件弯曲 智能化 神经网络 实时识别 实时预测

领　　域： [金属学及工艺] [金属学及工艺]