摘要
回弹是板料冲压成形中影响工件质量的重要因素 ,因为它是一个多变量相互作用的高度非线性问题 ,至今在解析和数值方法中未能找到一个很有效的解决途径。该文提出利用自适应模糊神经网络 (ANFIS)对非线性问题的良好逼近能力 ,采用基于有限元方法获得训练样本 ,经训练后得到具有回弹预测能力的ANFIS模型。实验验证了该方法的有效性。
Springback is a very important factor to effect the quality of sheet metal forming. Since it is a multi-variable coupled and high nonlinear problem, there hasn't been a effective either analytical or numerical approach proposed up to now . In this paper, a approach is proposed to predict the springback in sheet metal bending, which take the advantage of the excellent ablity of describing nonlinear problems of ANFIS . ANFIS is able to predict the springback with the training set obtained by using FEM software. Experiment verifys the effectivity of this approach.
出处
《计算机仿真》
CSCD
2003年第11期58-60,共3页
Computer Simulation
基金
福特--中国研究与发展基金 (5 0 12 2 15 4)
国家十五攻关项目 (2 0 0 2G5 3 12 )
